An autonomous obstacle avoidance control method for ship formation based on offset intervention guidance

Abstract

The present invention discloses an autonomous obstacle avoidance control method for a ship formation based on offset intervention guidance, comprising the following steps: obtaining a nonlinear mathematical model of a three-degree-of-freedom autonomous vehicle (ASV); obtaining a smooth reference path for a USV ship formation based on a virtual ship L1VS, and obtaining a virtual reference path for ship obstacle avoidance based on the obstacle avoidance virtual ship VS combined with a dynamic neural damping algorithm; designing a kinematic virtual control law for the USV ship formation based on the constructed offset intervention guidance mechanism; obtaining a robust kinematic error based on the kinematic error of the USV ship formation based on a constructed dynamic event triggering mechanism combined with robust neural damping technology; and designing an adaptive robust command filter controller based on the robust kinematic error to achieve autonomous obstacle avoidance control for the ship formation. This method solves the problem that traditional ship formation control methods cannot obtain information about moving and fixed obstacles in a timely manner when performing collision avoidance tasks in unknown waters in complex marine environments, and cannot effectively balance communication load and obstacle avoidance control accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous navigation applications of ships, and in particular to an autonomous obstacle avoidance control method for a ship formation based on offset intervention guidance. Background Art

[0002] Guidance and control are two important subsystems in path tracking control [3] In existing research results, guidance systems mostly rely on time to divide the path into straight segments and turning segments and calculate them separately. In addition, when performing simulated obstacle avoidance tasks, in most cases it is assumed that the position and motion information of mobile obstacles (such as ships, etc.) and fixed obstacles (such as islands and reefs, etc.) are known. In the control system, in order to reduce the problem of excessive communication load caused by continuous transmission of control commands, event triggering technology has been widely used in the field of path tracking control of USV formations, and it has good effects in terms of communication load. The dynamic surface control technology in the robust neural damping technology effectively solves the "computational explosion" problem and reduces the computational burden of the control system, but the generation of filtering errors is inevitable. Based on the above analysis, the existing USV formation collision avoidance and path tracking control algorithms still have the following two defects:

[0003] 1) For navigation situations with numerous waypoints (such as narrow waterways and island-reef waters), traditional LVS guidance algorithms can generate cumulative position and attitude errors at the switchover points (from straight segments to turning segments), which can affect the accuracy and safety of the guidance path. When performing collision avoidance missions in unknown waters, information about moving and fixed obstacles cannot be obtained in a timely manner, and existing collision avoidance algorithms such as artificial potential fields and virtual guidance methods do not perform optimally.

[0004] 2) Existing dynamic event-triggered technologies often dynamically adjust the trigger threshold based on system status. However, in complex marine environments, dynamic thresholds cannot serve as a single quantitative standard for triggering conditions. Furthermore, event-triggered technology can reduce USV control accuracy to a certain extent, and balancing communication load and control accuracy remains an open question. Summary of the Invention

[0005] The present invention provides an autonomous obstacle avoidance control method for a ship formation based on offset intervention guidance to overcome the above technical problems.

[0006] In order to achieve the above object, the technical solution of the present invention is:

[0007] A method for autonomous obstacle avoidance control of a ship formation based on offset intervention guidance specifically comprises the following steps:

[0008] S1: Obtain the nonlinear mathematical model of the three-degree-of-freedom ASV;

[0009] S2: Define the virtual ship L1VS and the obstacle avoidance virtual ship VS according to the nonlinear mathematical model;

[0010] Obtain a smooth reference path for the USV formation based on the virtual ship L1VS;

[0011] Based on the smooth reference path, the obstacle avoidance virtual ship VS is combined with the DWA algorithm to obtain the ship obstacle avoidance virtual reference path for the USV ship formation;

[0012] S3: Based on the constructed offset intervention guidance mechanism, the kinematic virtual control law of the USV formation is designed according to the virtual reference path of the ship obstacle avoidance;

[0013] S4: Based on the dynamic surface technology and the kinematic virtual control law, the kinematic error of the USV ship formation is obtained;

[0014] S5: Construct a dynamic event trigger mechanism to reduce the frequency of control command transmission between the ship controller and the actuator, and rewrite the kinematic error of the USV ship formation based on the dynamic event trigger mechanism to obtain the optimized kinematic error;

[0015] S6: Robust neural damping technology is used to robustify the optimization kinematic error to obtain a robust kinematic error;

[0016] S7: Design an adaptive robust command filter controller based on the robust kinematic error, and realize autonomous obstacle avoidance control of ship formation through the adaptive robust command filter controller.

[0017] The present invention provides an autonomous obstacle avoidance control method for a ship formation based on offset intervention guidance, which has the following two beneficial effects in terms of USV formation path tracking:

[0018] 1) The virtual reference path for obstacle avoidance of the USV formation is obtained by combining the obstacle avoidance virtual ship VS with the DWA algorithm. Based on the constructed offset intervention guidance mechanism, the kinematic virtual control law of the USV formation is designed according to the virtual reference path. This enables the unmanned underwater vehicles to achieve optimal collision avoidance according to the International Regulations for Preventing Collisions at Sea. This minimizes the rate at which the required signals are sent to the USV formation control module.

[0019] 2) Based on the dynamic event triggering of the output, a new empirical enhancement threshold rule is proposed to construct a dynamic event triggering mechanism that reduces the frequency of control command transmission between the ship controller and the actuator, ensuring a balance between communication efficiency and control accuracy. By designing an adaptive robust command filter controller, the control design is simplified and the filtering error is compensated, achieving precise control of autonomous obstacle avoidance for ship formations, which has good applicability in ship engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0021] Figure 1 This is a flow chart of the autonomous obstacle avoidance control method for a ship formation based on offset intervention guidance of the present invention;

[0022] Figure 2 This is the L1 virtual ship guidance principle diagram in this embodiment;

[0023] Figure 3 This is the overall path planning diagram for the autonomous obstacle avoidance of the ship formation in this embodiment;

[0024] Figure 4 This is a simulation diagram of the navigation trajectory of the unmanned ship formation in narrow waters in this embodiment;

[0025] Figure 5 The ship formation in this embodiment Input rudder angle and speed simulation diagram;

[0026] Figure 6 : is a simulation diagram of the position error and heading angle error of each ship in the formation in this embodiment;

[0027] Figure 7 This is a signal flow chart of the unmanned ship formation path tracking control system in this embodiment. DETAILED DESCRIPTION

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0029] This embodiment provides a method for autonomous obstacle avoidance control of a ship formation based on offset intervention guidance, such as Figure 1 As shown, the following steps are included:

[0030] S1: Obtain the nonlinear mathematical model of the three-degree-of-freedom ASV;

[0031] Specifically, the expression of the nonlinear mathematical model of the three-degree-of-freedom ASV is obtained as follows:

[0032] (1)

[0033] Where: Respectively represent the horizontal coordinate, vertical coordinate and yaw angle of the ship in the fixed coordinate system; Represent the speed variables of the ship's forward movement, drift and bow pitch respectively; express The first derivative of express The first derivative of represents the uncertainty term of the ship model; Represents the nonlinear terms in the ship's forward, drift and yaw directions; Indicates the time-varying external environmental interference to the ship in the forward, drift and pitch directions; and They represent the control inputs related to the ship's main engine speed and rudder angle respectively; and represents the time-varying ship actuator gain;

[0034] S2: Define the virtual ship L1VS and the obstacle avoidance virtual ship VS according to the nonlinear mathematical model;

[0035] Obtain a smooth reference path for the USV formation based on the virtual ship L1VS;

[0036] Specifically, the expression for obtaining the smooth reference path for the USV formation is:

[0037] (2)

[0038] Where: represents the desired position and heading angle of the virtual ship L1VS; and Represent the reference forward speed and bow angular velocity of the virtual ship L1VS respectively. Usually, is an artificially set constant. It is a variable calculated based on the geometric relationship between the virtual ship and the straight line connecting the waypoints; express The first derivative of; and for the virtual ship L1VS is in the straight line segment When the bow angular velocity ; For the virtual ship L1VS in the curve segment When the bow angular velocity And the turning acceleration at the waypoint ; Indicates connecting waypoints To waypoint line segment; Waypoints representing straight line segments; express Direction and The angle between directions; Figure 2 FIG shows a schematic diagram of the L1 guidance algorithm path planning of this embodiment. Figure 3 The collision avoidance mode is demonstrated using a formation of three USVs as an example; the red shaded area represents the obstacle area. The guidance method uses the concept of an inscribed circle, enabling the ship formation to navigate along a smooth reference trajectory during the turning phase. Figure 7 The principle structure of the algorithm proposed in this embodiment is shown. Among them, L1VS (L1 virtual ship, L1VS) and VS (VirtualShip, VS) do not consider any ship inertia and uncertainty factors, and the two virtual ships are ideal hulls without considering the effects of ship damping and inertia. The main task of the virtual ship L1VS is to Generate a smooth reference path for the entire fleet; the L1 guidance algorithm does not require computation time, thus eliminating the cumulative error generated during the turn to obtain a smooth reference path;

[0039] Based on the smooth reference path, the obstacle avoidance virtual ship VS is combined with the DWA algorithm to obtain the ship obstacle avoidance virtual reference path for the USV ship formation. The main task of the obstacle avoidance virtual ship in this embodiment is to generate an obstacle avoidance reference path for the USV by responding to dynamic obstacles (such as navigable ships) and sudden obstacle events (shipwrecks) in the environment in real time. In narrow waters, the USV formation needs to consider the constraints of the International Maritime Collision Prevention Regulations when avoiding navigable ships. When the target ship enters the range of the sensor, the encounter type is identified by calculating the relative orientation, as shown in Figure (2). The DWA algorithm calculates the speed sampling space, acceleration limit space and speed safety space in real time and takes the intersection to obtain the optimal set of speed sets.

[0040] The specific steps include:

[0041] S21: Define the desired position of the obstacle-avoiding virtual ship VS and and the expected velocity set and ;in Indicates the desired position abscissa, desired position ordinate, and desired pitch angle of the obstacle avoidance virtual ship VS; Indicates the expected forward speed and expected bow rolling speed of the obstacle avoidance virtual ship VS;

[0042] And the expected speed set The expression is

[0043] (3)

[0044] Where: Indicates the speed sampling space, acceleration limit space, and speed safety space obtained in real time based on the DWA algorithm; Indicates the minimum and maximum speed in the forward direction; Indicates the minimum and maximum speed in the heading direction; Indicates the current forward speed and yaw speed; Indicates the maximum surge acceleration and yaw acceleration; Indicates the distance from the current position to the nearest obstacle;

[0045] S22: Based on the smooth reference path according to the expected position and the expected velocity set , construct an evaluation function for improving the safety of USV ship formations and the accuracy of guidance signals;

[0046] This embodiment adds two evaluation functions on the basis of the traditional DWA algorithm to improve the safety of the formation and the accuracy of the guidance signal;

[0047] And the expression of the evaluation function is

[0048] (4)

[0049] (5)

[0050] Where: Represents the weight of the evaluation function; represents the heading evaluation function and ; Indicates the deviation between the heading angle and the path based on the waypoint; represents the safety distance function used to prioritize avoiding obstacles closest to the USV, and ; Represents the position coordinates of the obstacle closest to the candidate trajectory in the smooth reference path; Indicates the impact range of the obstacle; Represents the position coordinates of the candidate trajectory; represents the speed function used to ensure the safety of the braking distance under the current speed constraint; represents the collision avoidance steering function and ; represents a positive tuning parameter; Indicates the direction angle of the obstacle; represents the heading angle at the candidate trajectory; represents an intermediate variable and , ; express The vertical coordinate of the position of the smooth reference path at the moment The ordinate of the position of the smooth reference path at each moment; Represents time parameters; express The horizontal coordinate of the position of the smooth reference path at the moment The horizontal coordinate of the position of the smooth reference path at time ; represents the evaluation function; where By evaluating the function of the offset between the path generated by the obstacle avoidance virtual ship VS and the path based on the waypoints, it can avoid overshoot at turns and quickly converge to the path based on the waypoints after the collision avoidance action is completed. The smaller the value, the safer the path generated by the obstacle avoidance virtual ship VS is. The evaluation function is normalized and the ship guidance system selects the optimal speed set Generate real-time motion commands;

[0051] S23: Obtain the optimal speed set for ship guidance based on the evaluation function , and according to the optimal speed set Get the virtual reference path for USV formation to avoid ship obstacles, and its expression is:

[0052] (6)

[0053] Where: Indicates the horizontal and vertical coordinates of the virtual reference path for ship obstacle avoidance; Indicates the heading angle of the ship's obstacle avoidance virtual reference path; express The first derivative of express It is worth noting that due to the influence of the density of the sampling space, the obstacle avoidance virtual ship VS processes a large number of candidate paths during the autonomous selection process, making the generation of suboptimal trajectories inevitable, especially in the path tracking mode. It should remain stable. Next, this embodiment designs a kinematic virtual control law for realizing ship offset intervention guidance by constructing an offset intervention guidance mechanism.

[0054] S3: Based on the constructed offset intervention guidance mechanism, the kinematic virtual control law of the USV formation is designed according to the virtual reference path of the ship obstacle avoidance;

[0055] The specific steps include:

[0056] S31: Construct an offset intervention guidance mechanism based on the ship obstacle avoidance virtual reference path.

[0057] Conventional DWA algorithms generate transient stress on the servo actuator by continuously selecting the optimal signal at fixed sampling intervals, causing high-frequency heading angular velocity command switching. This mechanism can lead to actuator wear and increase closed-loop stability risks. Offset intervention guidance effectively mitigates these issues by reducing control input oscillations while maintaining trajectory tracking accuracy.

[0058] And the expression of the offset intervention guidance mechanism is

[0059] (7)

[0060] Where: represents a positive design parameter; express The variables in the evaluation function and Represents the error between adjacent sampling moments; represents the offset intervention guidance law of the USV; represents the sampling time; represents the trajectory error and ; represents the expected bow angle of the obstacle avoidance virtual ship VS and according to the positional relationship between the formation USV and the obstacle avoidance virtual ship VS, the guidance law of the USV Expressed as ; Indicates the horizontal and vertical coordinates of the USV's position;

[0061] S32: Design of kinematic virtual control law for USV formation based on offset intervention guidance mechanism;

[0062] And the expression of the kinematic virtual control law is

[0063] (8)

[0064] (9)

[0065] Where: Indicates the bow angle error of the USV formation; Indicates the The bow pitch angle of the USV; represents the position tracking error of the USV formation; express The first derivative of Indicates the USV ship positions The first derivative of represents the design parameters; The kinematic virtual control laws for the ship's forward and bowing directions are respectively represented;

[0066] S4: Based on the dynamic surface technology and the kinematic virtual control law, the kinematic error of the USV ship formation is obtained, which specifically includes the following steps:

[0067] S41: Based on the dynamic surface technology, a first-order low-pass filter is introduced, and the kinematic virtual control law is low-pass filtered through the first-order low-pass filter to obtain the dynamic surface signal;

[0068] In order to avoid the parameter explosion phenomenon in the derivation of the kinematic virtual control law, this embodiment introduces two first-order low-pass filters Perform low-pass filtering;

[0069] And the expression for obtaining dynamic surface signal is

[0070] (10)

[0071] Where: represents a time constant greater than zero; represents the Laplace operator; represents a dynamic surface signal and ; Represents the difference between the dynamic surface signal and the virtual control law, that is, ;

[0072] S42: The dynamic error of the USV ship formation is obtained based on the dynamic surface signal, and its expression is:

[0073] (11)

[0074] Where: They represent the dynamic errors of the USV formation’s forward direction and bow pitch direction respectively; In order to reduce the influence of filtering error on control accuracy, the compensation output error signal is defined as , and output error compensation signal is defined as ,in, represents a positive design parameter, Describe the input error compensation signal, where ;

[0075] S43: Based on the neural network approximation technology, the kinematic error of the USV ship formation is obtained by combining formula (10) and formula (1) according to the dynamic error;

[0076] And the expression of the kinematic error of the USV ship formation is

[0077] (12)

[0078] Where: Represents the kinematic error between the heading direction and the bow pitch direction of the USV formation; express The first derivative of Represents the neural network weight matrix; Represents the activation function of the neural network; represents the approximation error; represents transpose; Indicates the Control input of USV regarding main engine speed and rudder angle;

[0079] S5: Construct a dynamic event trigger mechanism to reduce the frequency of control command transmission between the ship controller and the actuator, and rewrite the kinematic error of the USV ship formation based on the dynamic event trigger mechanism to obtain the optimized kinematic error;

[0080] The specific steps include:

[0081] S51: Constructing a dynamic event triggering mechanism to reduce the frequency of control command transmission between the ship controller and the actuator. In order to reduce the frequency of control command transmission between the controller and the actuator, this embodiment proposes a new dynamic event triggering mechanism that combines navigation experience and output error;

[0082] And the expression of the dynamic event triggering mechanism is

[0083] ,

[0084] (13)

[0085] Where: Respectively represent USV ships Time and ship main engine speed and rudder angle Related control inputs; Represents time parameters; and Respectively represent the ship's main engine speed and rudder angle The trigger time parameter; Indicates the USV ships and ship main engine speed and rudder angle The control input error is related to ; Indicates the The expected control input of each USV; Indicates the Actual control input of USV ships; represents the designed threshold parameter; Indicates the trigger threshold; Represents the position tracking error Design parameters;

[0086] S52: Simplify the dynamic event triggering mechanism to obtain the simplified dynamic event triggering mechanism:

[0087] (14)

[0088] Where: express The abbreviation of , ; express The abbreviation of Indicates the time trigger boundary parameter item of the design;

[0089] S53: Substitute formula (14) into formula (12) to rewrite the kinematic error of the USV ship formation to obtain the optimized kinematic error;

[0090] And the expression of the optimized kinematic error is

[0091] (15).

[0092] S6: Robust neural damping technology is used to robustly process the optimized kinematic error to obtain a robust kinematic error, which specifically includes the following steps:

[0093] S61: Robust neural damping technology is used to robustly handle nonlinear terms, event-triggered parameter terms, model uncertainties, and real-time external disturbances in the optimization kinematics error.

[0094] And the expression for robust processing is

[0095] (16)

[0096] Where: represents the design parameters and

[0097] ,

[0098] ; represents the robust neural damping term, i.e.

[0099] , ; Indicates the upper bound of real-time external disturbance; represents a positive constant; express The upper bound of

[0100] S62: Based on formula (16), the optimized kinematic error is improved to obtain the robust kinematic error;

[0101] And the expression for optimizing kinematic error is rewritten as

[0102] (17)

[0103] S7: Design an adaptive robust command filter controller based on robust kinematic errors, and implement autonomous obstacle avoidance control for ship formations through the adaptive robust command filter controller;

[0104] In a specific embodiment, a method for designing an adaptive robust command filter controller specifically includes the following steps:

[0105] S71: Defining intermediate control input variables based on robust kinematic errors Compensation input tracking error signal The gain coefficient of the control input in formula (17) is the actuator gain function coupled with the event trigger parameter. It is not easy to obtain directly in engineering practice, so it is necessary to adaptively process its uncertain information, that is, to introduce the variable As The estimated value of further defines the intermediate control input variable ;

[0106] And the expression of the intermediate control input variable is

[0107] (18)

[0108] Where: Indicates intermediate parameters;

[0109] The expression of the compensated input tracking error signal is:

[0110] (19)

[0111] (20)

[0112] Where: Indicates the input tracking error compensation amount; represents a positive controller design parameter; represents a positive design parameter and Initial value of ; express The first derivative of

[0113] S72: Design an adaptive robust command filter controller based on the intermediate control input variable and the compensation input tracking error signal, and the expression of the adaptive robust command filter controller is:

[0114] (twenty one)

[0115] (twenty two)

[0116] Where: represents a positive gain adaptation design parameter; represents the design parameters of the controller; represents the adaptive design parameters of the robust neural damping estimator; represents the input of the adaptive robust command filter controller; express estimated value of; express The initial value of express The first-order derivative of is the adaptive law of the robust neural damping estimator designed to stabilize the closed-loop control system.

[0117] In order to verify the effectiveness of the method proposed in this embodiment, a formation consisting of three USVs is used as the controlled object, where each ship is an underactuated ship with a length of 38m and a displacement of 1.18*105kg. By designing a digital simulation experiment of the narrow water navigation task of the ship formation in the marine environment, the planned route consists of 4 waypoints. ; The initial motion state of the corresponding unmanned ship formation is:

[0118] ,

[0119] ,

[0120] ,

[0121] The expected path tracking speed is , Figures 4 to 7 The simulation results of ship-UAV collaborative search under level 4 sea conditions are respectively presented on the MATLAB simulation platform; Figure 4 The trajectory diagram of the USV formation performing collision avoidance mission in narrow waters. Figure 4 It can be seen that this embodiment is applicable to three common encounter situations, emergency collision avoidance and inter-ship collision avoidance, and demonstrates good effects. Figure 5 Indicates the formation system The local enlarged image shows the superiority of the algorithm. It is obvious that the actual input (red dotted line) has a lower frequency of change in the input command than the control input (blue solid line), indicating that the method proposed in this embodiment is effective in reducing the computational load and actuator wear. Figure 6 It represents the position error and attitude error of each ship in the formation. It is obvious that the error is kept within a reasonable range.

[0122] The method described in this embodiment has the following two beneficial effects in USV formation path tracking:

[0123] 1) The virtual reference path for obstacle avoidance of the USV formation is obtained by combining the obstacle avoidance virtual ship VS with the DWA algorithm. Based on the constructed offset intervention guidance mechanism, the kinematic virtual control law of the USV formation is designed according to the virtual reference path. This enables the unmanned underwater vehicles to achieve optimal collision avoidance according to the International Regulations for Preventing Collisions at Sea. This minimizes the rate at which the required signals are sent to the USV formation control module.

[0124] 2) Based on the dynamic event triggering of the output, a new empirical enhancement threshold rule is proposed to construct a dynamic event triggering mechanism that reduces the frequency of control command transmission between the ship controller and the actuator, ensuring a balance between communication efficiency and control accuracy. By designing an adaptive robust command filter controller, the control design is simplified and the filtering error is compensated, achieving precise control of autonomous obstacle avoidance for ship formations, which has good applicability in ship engineering.

[0125] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for autonomous obstacle avoidance control of a ship formation based on offset intervention guidance, characterized in that: The specific steps include: S1: Obtain the nonlinear mathematical model of the three-degree-of-freedom USV; S2: Define the virtual ship L1VS and the obstacle avoidance virtual ship VS according to the nonlinear mathematical model; Obtain a smooth reference path for the USV formation based on the virtual ship L1VS; Based on the smooth reference path, the obstacle avoidance virtual ship VS is combined with the DWA algorithm to obtain the ship obstacle avoidance virtual reference path for the USV ship formation, including: S21: Define the desired position of the obstacle-avoiding virtual ship VS and and the expected speed set and ;in Indicates the desired position abscissa, desired position ordinate, and desired pitch angle of the obstacle avoidance virtual ship VS; Indicates the expected forward speed and expected bow rolling speed of the obstacle avoidance virtual ship VS; And the expected speed set The expression is (3) Where: Indicates the speed sampling space, acceleration limit space, and speed safety space obtained in real time based on the DWA algorithm; Indicates the minimum and maximum speed in the forward direction; Indicates the minimum and maximum speed in the heading direction; Indicates the current forward speed and yaw speed; Indicates the maximum surge acceleration and yaw acceleration; Indicates the distance from the current position to the nearest obstacle; S22: Based on the smooth reference path according to the expected position and the expected speed set , construct an evaluation function for improving the safety of USV ship formations and the accuracy of guidance signals; And the expression of the evaluation function is (4) (5) Where: Represents the weight of the evaluation function; represents the heading evaluation function and ; Indicates the deviation between the heading angle and the path based on the waypoint; represents the safety distance function used to prioritize avoiding obstacles closest to the USV, and ; Represents the position coordinates of the obstacle closest to the candidate trajectory in the smooth reference path; Indicates the impact range of the obstacle; Represents the position coordinates of the candidate trajectory; represents the speed function used to ensure the safety of the braking distance under the current speed constraint; represents the collision avoidance steering function and ; represents a positive tuning parameter; Indicates the direction angle of the obstacle; represents the heading angle at the candidate trajectory; represents an intermediate variable and , ; express The vertical coordinate of the position of the smooth reference path at the moment The ordinate of the position of the smooth reference path at each moment; Represents time parameters; express The horizontal coordinate of the position of the smooth reference path at the moment The horizontal coordinate of the position of the smooth reference path at time ; represents the evaluation function; S23: Obtain the optimal speed set for ship guidance based on the evaluation function , and according to the optimal speed set Get the virtual reference path for USV formation to avoid ship obstacles, and its expression is: (6) Where: Indicates the horizontal and vertical coordinates of the virtual reference path for ship obstacle avoidance; Indicates the heading angle of the ship's obstacle avoidance virtual reference path; express The first derivative of express The first derivative of S3: Based on the constructed offset intervention guidance mechanism, the kinematic virtual control law of the USV formation is designed according to the virtual reference path of the ship obstacle avoidance; S4: Based on the dynamic surface technology and the kinematic virtual control law, the kinematic error of the USV ship formation is obtained; S5: Construct a dynamic event trigger mechanism to reduce the frequency of control command transmission between the ship controller and the actuator, and rewrite the kinematic error of the USV ship formation based on the dynamic event trigger mechanism to obtain the optimized kinematic error; S6: Robust neural damping technology is used to robustify the optimization kinematic error to obtain a robust kinematic error; S7: Design an adaptive robust command filter controller based on the robust kinematic error, and realize autonomous obstacle avoidance control of ship formation through the adaptive robust command filter controller.

2. The method for autonomous obstacle avoidance control of a ship formation based on offset intervention guidance according to claim 1, characterized in that: The nonlinear mathematical model of the three-degree-of-freedom ASV obtained in S1 is expressed as (1) Where: Respectively represent the horizontal coordinate, vertical coordinate and yaw angle of the ship in the fixed coordinate system; Represent the speed variables of the ship's forward movement, drift and bow pitch respectively; express The first derivative of express The first derivative of represents the uncertainty term of the ship model; Represents the nonlinear terms in the ship's forward, drift and yaw directions; Indicates the time-varying external environmental interference to the ship in the forward, drift and pitch directions; and They represent the control inputs related to the ship's main engine speed and rudder angle respectively; and represents the time-varying ship actuator gain.

3. The method for autonomous obstacle avoidance control of a ship formation based on offset intervention guidance according to claim 2 is characterized in that: The smooth reference path for the USV formation is obtained in S2, and its expression is: (2) Where: represents the desired position and heading angle of the virtual ship L1VS; and represent the reference forward speed and bow angular velocity of the virtual ship L1VS respectively; express The first derivative of; and for the virtual ship L1VS is in the straight line segment When the bow angular velocity ; For the virtual ship L1VS in the curve section When the bow angular velocity And the turning acceleration at the waypoint ; Indicates connecting waypoints To waypoint line segment; Waypoints representing straight line segments; express Direction and The angle of direction.

4. The method for autonomous obstacle avoidance control of a ship formation based on offset intervention guidance according to claim 3 is characterized in that: The S3 specifically includes the following steps: S31: Construct an offset intervention guidance mechanism based on the ship obstacle avoidance virtual reference path. And the expression of the offset intervention guidance mechanism is (7) Where: represents a positive design parameter; express The variables in the evaluation function and Represents the error between adjacent sampling moments; represents the offset intervention guidance law of the USV; represents the sampling time; represents the trajectory error and ; represents the expected pitch angle of the obstacle-avoiding virtual ship VS and ; Indicates the horizontal and vertical coordinates of the USV's position; S32: Design of kinematic virtual control law for USV formation based on offset intervention guidance mechanism; And the expression of the kinematic virtual control law is (8) (9) Where: Indicates the bow angle error of the USV formation; Indicates the The bow pitch angle of the USV; represents the position tracking error of the USV fleet; express The first derivative of Indicates the USV ship positions The first derivative of represents the design parameters; They represent the kinematic virtual control laws for the ship's forward direction and bow pitch direction respectively.

5. The method for autonomous obstacle avoidance control of a ship formation based on offset intervention guidance according to claim 4 is characterized in that: The S4 specifically includes the following steps: S41: Based on the dynamic surface technology, a first-order low-pass filter is introduced, and the kinematic virtual control law is low-pass filtered through the first-order low-pass filter to obtain the dynamic surface signal; And the expression for obtaining dynamic surface signal is (10) Where: represents a time constant greater than zero; represents the Laplace operator; represents a dynamic surface signal and ; Represents the difference between the dynamic surface signal and the virtual control law, that is, ; S42: The dynamic error of the USV ship formation is obtained based on the dynamic surface signal, and its expression is: (11) Where: They represent the dynamic errors of the USV formation’s forward direction and bow pitch direction respectively; Indicates the The forward speed and bow pitch speed of the USV; S43: Based on the neural network approximation technology, the kinematic error of the USV ship formation is obtained by combining formula (10) and formula (1) according to the dynamic error; And the expression of the kinematic error of the USV ship formation is (12) Where: Represents the kinematic error between the heading direction and the bow pitch direction of the USV formation; express The first derivative of Represents the neural network weight matrix; Represents the activation function of the neural network; represents the approximation error; represents transpose; Indicates the The USV's control inputs are about the ship's main engine speed and rudder angle.

6. The method for autonomous obstacle avoidance control of a ship formation based on offset intervention guidance according to claim 5, characterized in that: The S5 specifically includes the following steps: S51: Construct a dynamic event triggering mechanism to reduce the frequency of control command transmission between ship controllers and actuators; And the expression of the dynamic event triggering mechanism is , (13) Where: Respectively represent USV ships Time and ship main engine speed and rudder angle Related control inputs; Represents time parameters; and Respectively represent the ship's main engine speed and rudder angle The trigger time parameter; Indicates the USV ships and ship main engine speed and rudder angle The control input error is related to ; Indicates the The expected control input of each USV; Indicates the Actual control input of USV ships; represents the designed threshold parameter; Indicates the trigger threshold; Represents the position tracking error Design parameters; S52: Simplify the dynamic event triggering mechanism to obtain the simplified dynamic event triggering mechanism: (14) Where: express The abbreviation of , ; express The abbreviation of Indicates the time trigger boundary parameter item of the design; S53: Substitute formula (14) into formula (12) to rewrite the kinematic error of the USV ship formation to obtain the optimized kinematic error; And the expression of the optimized kinematic error is (15)。 7. The method for autonomous obstacle avoidance control of a ship formation based on offset intervention guidance according to claim 6, characterized in that: The S6 specifically includes the following steps: S61: Robust neural damping technology is used to robustly handle nonlinear terms, event-triggered parameter terms, model uncertainties, and real-time external disturbances in the optimization kinematics error. And the expression for robust processing is (16) Where: represents the design parameters and ， ; represents the robust neural damping term, i.e. , ; Indicates the upper bound of real-time external disturbance; represents a positive constant; express The upper bound of S62: Based on formula (16), the optimized kinematic error is rewritten to obtain the robust kinematic error; And the expression for optimizing kinematic error is rewritten as (17)。 8. The method for autonomous obstacle avoidance control of a ship formation based on offset intervention guidance according to claim 7 is characterized in that: The method for designing an adaptive robust command filter controller based on robust kinematic errors in S7 specifically comprises the following steps: S71: Defining intermediate control input variables based on robust kinematic errors Compensation input tracking error signal ; And the expression of the intermediate control input variable is (18) Where: Indicates intermediate parameters; The expression of the compensated input tracking error signal is: (19) (20) Where: Indicates the input tracking error compensation amount; represents a positive controller design parameter; represents a positive design parameter and Initial value of ; express The first derivative of S72: Design an adaptive robust command filter controller based on the intermediate control input variable and the compensation input tracking error signal, and the expression of the adaptive robust command filter controller is: (21) (22) Where: represents a positive gain adaptation design parameter; represents the design parameters of the controller; represents the adaptive design parameters of the robust neural damping estimator; represents the input of the adaptive robust command filter controller; express estimated value of; express The initial value of express The first derivative of .

Images