An underactuated unmanned surface vessel formation control method based on output redefinition

Abstract

A method for underactuated unmanned surface vessel (USV) formation control based on output redefinition is disclosed, relating to the field of underactuated USV formation control technology. To address the technical shortcomings of existing underactuated USV formation control technologies, such as insufficient control input, model uncertainty and external disturbances, and the differential explosion problem in traditional control methods, this invention provides a method for underactuated USV formation control based on output redefinition. The method includes: establishing a kinematic and dynamic model of the underactuated USV and decomposing it into an attitude subsystem and a position subsystem; designing attitude controllers and position controllers for the attitude and position subsystems respectively to approximate the system's nonlinear terms; avoiding repeated differentiation of the virtual control law in the model using a first-order filter; and combining an adaptive control law based on the minimum learning parameter method and an auxiliary adaptive law. This method is suitable for application in underactuated USV formation control.

Description

Technical Field

[0001] This relates to the field of underactuated unmanned surface vessel (USV) formation control technology, specifically to underactuated USV formation control based on output redefinition. Background Technology

[0002] In recent years, with the rapid development of unmanned driving technology and marine engineering technology, research on unmanned surface vehicles (USVs) has gradually become a focus, especially in fields such as military, search and rescue, and environmental monitoring, where the importance of formation control technology is increasingly prominent. USV formations can achieve coordinated cooperation across multiple objectives and tasks, offering greater flexibility, better mission adaptability, and higher operational efficiency compared to single USVs. In numerous scenarios such as marine search and rescue, environmental monitoring, resource exploration, navigation guidance, and military escort, the use of USV formations can significantly improve the coverage and reliability of mission execution.

[0003] Existing unmanned surface vessel (USV) formation control methods include fuzzy control, backstepping control, sliding mode control, and graph theory-based control algorithms. These methods are widely used in fully actuated USV formations and are relatively mature technologies. Fully actuated USVs can achieve independent control of all degrees of freedom through control inputs, thus their formation control problem is relatively simple. However, with the increasing complexity of mission requirements, the use of underactuated unmanned surface vehicles (UUSVs) is gradually increasing. Because underactuated USVs have fewer control inputs than system degrees of freedom, they face the problem of insufficient control degrees of freedom during navigation, especially in complex environments, where their stability and accuracy are difficult to guarantee. Therefore, formation control methods for underactuated USVs are currently a key research focus and challenge.

[0004] In existing research, some scholars have proposed combining sliding mode control with neural networks to improve the robustness of control systems. However, when applied to underactuated unmanned surface vessel (USV) formations, these methods still suffer from problems such as insufficient control input, model uncertainty, and external disturbances, making it difficult to guarantee formation stability. Furthermore, while backstepping control can address the insufficient degrees of freedom to some extent, its high system complexity often leads to the "differential explosion" problem in differential calculations, significantly limiting its feasibility and real-time performance in practical applications. On the other hand, radial basis function (RBF) neural networks are used to approximate nonlinear terms in the system; however, due to the large number of learning parameters, the computational complexity in real-time control remains high, affecting the controller's fast response capability.

[0005] Furthermore, existing formation control methods generally lack effective interference suppression measures. Since unmanned surface vessels (USVs) navigate in the marine environment, they are affected by external disturbances such as waves and wind speed. Effectively suppressing the impact of these disturbances on formation control becomes a key challenge in achieving stable formation. Although some methods introduce adaptive control laws, these often require a large number of learning parameters, leading to complex controller design and difficulty in achieving real-time adjustment.

[0006] In summary, existing underactuated unmanned surface vessel (USV) formation control technologies suffer from the following main problems: insufficient control input, making it difficult to achieve independent control of all degrees of freedom; model uncertainty and external disturbances affect formation stability; and traditional control methods suffer from differential explosion problems when dealing with complex nonlinear systems, resulting in high computational complexity and difficulty in meeting the requirements of real-time control.

[0007] Therefore, there is an urgent need for an underactuated unmanned surface vessel (USV) formation control method that can reduce computational complexity and enhance the system's anti-interference capability, so as to achieve more stable and efficient formation control. Summary of the Invention

[0008] To address the shortcomings of existing underactuated unmanned surface vessel (USV) formation control technologies, such as insufficient control input, model uncertainty, external disturbances, and the differential explosion problem in traditional control methods when dealing with complex nonlinear systems, the present invention provides the following technical solution:

[0009] A method for formation control of underactuated unmanned surface vessels based on output redefinition, the method comprising:

[0010] The steps to establish the kinematic and dynamic models of the underactuated unmanned surface vessel and decompose them into attitude subsystems and position subsystems;

[0011] The steps involve designing attitude controllers and position controllers respectively through the attitude subsystem and position subsystem to approximate the nonlinear terms of the system.

[0012] The step of avoiding repeated differentiation of the virtual control law in the model by using a first-order filter;

[0013] The steps combine the adaptive control law and the auxiliary adaptive law based on the minimum learning parameter method.

[0014] Furthermore, a preferred embodiment is provided, in which the kinematics and dynamics model of the underactuated unmanned surface vessel is redefined and decomposed into an attitude subsystem and a position subsystem.

[0015] Furthermore, a preferred implementation is provided, which combines the minimum learning parameter method with a radial basis neural network to approximate the nonlinear terms of the system.

[0016] Furthermore, a preferred implementation method is provided, which uses a dynamic surface control method to avoid repeated differentiation of the virtual control law in the model through a first-order filter.

[0017] Furthermore, a preferred embodiment is provided, which also includes the step of performing stability analysis on the attitude controller and position controller in the control system of the underactuated unmanned surface vessel by constructing a Lyapunov function.

[0018] Furthermore, a preferred embodiment is provided, which further includes the step of incorporating a fuzzy control strategy into the attitude controller and the position controller.

[0019] Based on the same inventive concept, the present invention also provides an underactuated unmanned surface vessel (USV) formation control device based on output redefinition, the device comprising:

[0020] Establish the kinematic and dynamic models of the underactuated unmanned surface vessel and decompose them into attitude subsystems and position subsystems.

[0021] The attitude subsystem and position subsystem are used to design attitude controllers and position controllers respectively, which are modules that approximate the nonlinear terms of the system.

[0022] A module that avoids repeated differentiation of the virtual control law in the model through a first-order filter;

[0023] A module combining adaptive control law and auxiliary adaptive law based on the minimum learning parameter method.

[0024] Based on the same inventive concept, the present invention also provides a computer storage medium for storing a computer program, wherein when the computer reads the computer program, the computer executes the method described thereon.

[0025] Based on the same inventive concept, the present invention also provides a computer, including a processor and a storage medium, wherein when the processor reads a computer program stored in the storage medium, the computer executes the method described thereon.

[0026] Based on the same inventive concept, the present invention also provides a computer program product, which, when executed, implements the method described.

[0027] Compared with the prior art, the advantages of the technical solution provided by the present invention are as follows:

[0028] By redefining the output, the original model of the underactuated unmanned surface vessel (USV) is decomposed into attitude and position subsystems, greatly simplifying the complexity of the overall control problem. This allows the controller design to more effectively achieve independent control of attitude and position. Compared with existing technologies, this approach effectively solves the problem of insufficient control input, making the degree-of-freedom control of the underactuated system more precise and significantly improving the stability and robustness of formation control.

[0029] The design of the attitude and position controllers, combining the minimum learning parameter method with radial basis function neural networks, significantly improves the system's nonlinear approximation capability and robustness. Compared with traditional single control methods, this combined approach demonstrates stronger adaptability to model uncertainties and external disturbances, while reducing computational complexity and improving real-time performance by decreasing the number of learning parameters.

[0030] The introduction of dynamic surface control technology avoids the "differential explosion" problem during control law design. By using a first-order filter to replace repetitive differential operations, the computational load is greatly reduced. This makes the present invention more suitable for real-time applications compared to traditional backstepping control, especially in complex marine environments, enabling faster and more accurate control responses.

[0031] By combining an adaptive control law based on the minimum learning parameter method, the system can dynamically adjust control parameters during task execution, enhancing its anti-interference capability. Compared to existing fixed-parameter control methods, this adaptability enables the unmanned surface vessel to better maintain formation stability and consistency when facing changes and disturbances in the external environment, significantly improving control performance.

[0032] Through stability analysis and the construction of Lyapunov functions, it is proven that the designed control law enables the system error to converge to a small region near zero, and all signals satisfy the globally eventually uniform bounded property (SGUUB). Compared with traditional methods, this process provides a rigorous mathematical proof, ensuring the stability and effectiveness of the controller design and further enhancing the reliability of the system.

[0033] Simulation experiments verified the effectiveness of the designed method, demonstrating that six unmanned surface vessels successfully transitioned from their initial state to the desired formation, with all control signals exhibiting continuity and fluctuation-free operation. Compared to other methods, the controller designed in this invention better maintains the smoothness of control signals, reduces system fluctuation, and thus improves the control quality of the formation and the reliability of mission execution.

[0034] Suitable for use in the formation control of underactuated unmanned surface vessels. Attached Figure Description

[0035] Figure 1This is a schematic diagram of the formation trajectory on a plane.

[0036] Figure 2 A schematic diagram of the model error variables for 6 ships;

[0037] Figure 3 The control input signal τ for the vehicle X and τ σ Schematic diagram;

[0038] Figure 4 This is a schematic diagram of the virtual control laws α1 and α2 of the Vehicle. Detailed Implementation

[0039] To make the advantages and benefits of the technical solution provided by the present invention clearer, the technical solution provided by the present invention will now be described in further detail with reference to the accompanying drawings, specifically:

[0040] Implementation Method 1: This implementation method provides an underactuated unmanned surface vessel (USV) formation control method based on output redefinition. The method includes:

[0041] The steps to establish the kinematic and dynamic models of the underactuated unmanned surface vessel and decompose them into attitude subsystems and position subsystems;

[0042] The steps involve designing attitude controllers and position controllers respectively through the attitude subsystem and position subsystem to approximate the nonlinear terms of the system.

[0043] The step of avoiding repeated differentiation of the virtual control law in the model by using a first-order filter;

[0044] The steps combine the adaptive control law and the auxiliary adaptive law based on the minimum learning parameter method.

[0045] Specifically, including:

[0046] 1. Establish kinematic and dynamic models and redefine the output.

[0047] First, the kinematic and dynamic model of the underactuated unmanned surface vessel (UUSV) is established, and its output is redefined. By redefining the output, the original model is decomposed into attitude and position subsystems. This transformation simplifies the overall control problem into more manageable components, laying the foundation for subsequent controller design.

[0048] 2. Design the attitude controller and position controller.

[0049] After redefining the model, attitude controllers and position controllers were designed. The attitude controllers are mainly used to control the attitude of each unmanned surface vessel (USV) to achieve the target direction and angle; the position controllers are used to control the position of the USVs to achieve the expected formation shape.

[0050] Specifically, through the attitude subsystem, error variables are defined and a virtual control law is constructed. Then, the minimum learning parameter method (MLP) combined with radial basis function (RBF) is used to approximate the nonlinear terms in the system. Next, the control law for the position subsystem is designed to further ensure precise position control.

[0051] 3. Introduction of Dynamic Surface Control (DSC) technology

[0052] In the process of controller design, Dynamic Surface Control (DSC) technology is introduced to solve the "differential explosion" problem in control design. By introducing a first-order filter, DSC technology avoids repeated differentiation of the virtual control law, greatly reducing computational complexity and enhancing the real-time performance of the control system.

[0053] 4. Combining adaptive control law with disturbance suppression

[0054] To enhance the system's anti-interference capability, this invention incorporates an adaptive control law based on the minimum learning parameter method. Through this adaptive law, the system can continuously adjust control parameters during task execution to cope with changes in the external environment and model uncertainties. Simultaneously, the design of an auxiliary adaptive law further improves the system's stability.

[0055] 5. Stability Analysis

[0056] After the controller design was completed, a stability analysis was performed on the entire formation system. First, a Lyapunov function was constructed for the attitude controller, and it was proven that the designed control law enables the attitude error to converge to a small region near zero, and that all signals within the system satisfy the globally eventually uniform bounded property (SGUUB). Next, a similar stability analysis was performed on the position controller to ensure that the position error also converges to the desired range.

[0057] 6. Simulation Verification

[0058] Finally, simulation experiments were conducted to verify the effectiveness of the invention. Simulation results show that, utilizing the designed control and adaptive laws, the six unmanned surface vessels successfully achieved the desired formation from the initial state, realizing the transition from a wedge formation to a column formation and then to a line formation. All control signals exhibited continuous operation without chattering, verifying the superiority and practicality of the proposed method.

[0059] Implementation Method 2: This implementation method further defines the underactuated unmanned surface vessel (USV) formation control method based on output redefinition provided in Implementation Method 1. It decomposes the kinematic and dynamic models of the USV into attitude subsystems and position subsystems by redefining the output.

[0060] Implementation Method 3: This implementation method further defines the underactuated unmanned surface vessel formation control method based on output redefinition provided in Implementation Method 1. It combines the minimum learning parameter method with radial basis neural networks to approximate the nonlinear terms of the system.

[0061] Implementation Method 4: This implementation method further defines the underactuated unmanned surface vessel formation control method based on output redefinition provided in Implementation Method 1. Based on the dynamic surface control method, it avoids the repeated differentiation of the virtual control law in the model by using a first-order filter.

[0062] Implementation Method 5: This implementation method further defines the underactuated unmanned surface vessel (USV) formation control method based on output redefinition provided in Implementation Method 1. It also includes the step of performing stability analysis on the attitude controller and position controller in the underactuated USV control system by constructing a Lyapunov function.

[0063] Implementation Method Six: This implementation method further defines the underactuated unmanned surface vessel formation control method based on output redefinition provided in Implementation Method One, and also includes the step of incorporating a fuzzy control strategy in the attitude controller and position controller.

[0064] Implementation Method Seven: This implementation method provides an underactuated unmanned surface vessel (USV) formation control device based on output redefinition. The device includes:

[0065] Establish the kinematic and dynamic models of the underactuated unmanned surface vessel and decompose them into attitude subsystems and position subsystems.

[0066] The attitude subsystem and position subsystem are used to design attitude controllers and position controllers respectively, which are modules that approximate the nonlinear terms of the system.

[0067] A module that avoids repeated differentiation of the virtual control law in the model through a first-order filter;

[0068] A module combining adaptive control law and auxiliary adaptive law based on the minimum learning parameter method.

[0069] Implementation Method 8: This implementation method provides a computer storage medium for storing computer programs. When the computer reads the computer program, the computer executes the method provided in Implementation Method 1.

[0070] Implementation Method Nine: This implementation method provides a computer, including a processor and a storage medium. When the processor reads a computer program stored in the storage medium, the computer executes the method provided in Implementation Method One.

[0071] Implementation Method 10: This implementation method provides a computer program product. As a computer program, when the computer program is executed, it implements the method provided in Implementation Method 1.

[0072] Implementation Method Eleven: Combination Figure 1-4 This embodiment describes the technical solution provided above in further detail through specific examples. Specifically:

[0073] include:

[0074] (1) Redefine the output of the kinematic and dynamic model of UUSV and decompose it into attitude subsystem and position subsystem;

[0075] (2) Design the attitude controller and position controller;

[0076] (3) Stability analysis;

[0077] 1. In step 1, the kinematics and dynamics of the UUSV are transformed and redefined, decomposed into an attitude subsystem and a position subsystem, as shown below:

[0078] Considering a cluster of N UUSVs, the kinematic and dynamic equations of the UUSVs are described as follows:

[0079]

[0080]

[0081] Where i = 1, 2, ..., N represents the UUSV number, P i =[x i ,y i ] T Let (x) be the position vector of UUSV in the geodetic coordinate system. i ,y i () represents the position coordinates; Let θ be the rotation matrix related to the bow roll angle. i For bow roll angle; Q i =[u i ,v i ] T Let u be the velocity vector in the ship's coordinate system. i and v i These represent forward and sideways speeds, respectively, r i This represents the bow roll rate of the UUSV. F i (x i ,y i )=[f i (x i ),f i (y i )]T and f i (σ i () represents an unknown model with unmodeled dynamics; as well as External interference during UUSV navigation as well as This is the thrust output of the UUSV.

[0082] Since the above model is an underactuated model, for the convenience of further research, the UUSV model can be converted into an attitude subsystem and a position subsystem, as shown below:

[0083]

[0084] in and This represents the state-related nonlinear terms that include dynamics not modeled by the model and external disturbances.

[0085] 2. In step 2, design the attitude controller and position controller as follows:

[0086] The distributed controller design for the UUSV formation consists of two parts. The first part involves designing control laws to control the attitude of each UUSV through the attitude subsystem. Then, a control law that can control the position of each UUSV is designed using the position subsystem. The specific design process is as follows.

[0087] (1) For the attitude subsystem, define the error variable z. i1 ,z i2 ,z i3 as follows

[0088]

[0089] Where α i1 For virtual control laws, For filtering control law, Each element in is α i1 The time constant γ of this filter is obtained through a first-order filter. i Given by the corresponding diagonal elements, it is a diagonal positive definite matrix.

[0090] First, construct the virtual control law as follows:

[0091]

[0092] Where K i1 ,K i0 ∈R 3×3 All are diagonal matrices with positive diagonal elements and 0 for the rest, and ε is a variable positive constant.

[0093] Next, we will design the attitude control law for each UUSV.

[0094]

[0095] exist Zhong K i2 It is a diagonal matrix, where the diagonal elements are positive constants and the other elements are 0.

[0096] MLP-based adaptive control law and auxiliary adaptive law as follows

[0097]

[0098] in φ i >0, This represents an estimate of φ. The initial value is This represents the estimation error. Γ > 0 represents the adaptive gain. For feedback gain. It is for b di =ε N +b d The estimate, To estimate the error, Γ b For positive adaptive gain, This is a positive feedback gain.

[0099] (2) For the position subsystem, define the error variable z. i4 ,z i5 ,z i6 as follows

[0100]

[0101] Where α i4 For virtual control laws, For filtering control law, Each element in is α i4 The time constant χ of this filter is obtained through a first-order filter. i Given by the corresponding diagonal elements, it is a diagonal positive definite matrix.

[0102] Constructing virtual control laws as

[0103]

[0104] Where M i1 M i0 ∈R 3×3All are diagonal matrices with positive diagonal elements and 0 for the rest of the elements, and δ is a variable positive constant.

[0105] Next, we will design the position control law for each UUSV.

[0106]

[0107] exist M i2 It is a diagonal matrix, where the diagonal elements are positive constants and the other elements are 0.

[0108] MLP-based adaptive control law and auxiliary adaptive law as follows

[0109]

[0110] in Represents φ X The estimate, The initial value is To estimate the error. Γ X >0 indicates adaptive gain. For feedback gain. Yes The estimate, To estimate the error, For positive adaptive gain, This is a positive feedback gain.

[0111] 3. Perform stability analysis in step 3, as shown below:

[0112] (1) For the attitude control law, first define the Lyapunov function V1:

[0113]

[0114] It can be concluded that:

[0115]

[0116] Next, define the Lyapunov function V2:

[0117]

[0118] It can be concluded that:

[0119]

[0120] Further define the Lyapunov function V3:

[0121]

[0122] Ultimately, we can obtain:

[0123]

[0124] in:

[0125]

[0126] Therefore, the designed control law and adaptive law can make the attitude error of each UUSV in the formation converge to a small region near zero, and all signals in the system satisfy the global eventually bounded property.

[0127] (2) For the position control law, first define the Lyapunov function V4:

[0128]

[0129] It can be concluded that:

[0130]

[0131] Next, define the Lyapunov function V5:

[0132]

[0133] It can be concluded that:

[0134]

[0135] Further define the Lyapunov function V6:

[0136]

[0137] Ultimately, we can obtain:

[0138]

[0139] in:

[0140]

[0141] Therefore, the designed control law and adaptive law can make the position error of each UUSV in the formation converge to a small region near zero, and all signals in the system satisfy the global eventually consistent boundedness.

[0142] Compared with the prior art, the beneficial effects of this implementation are: the dynamics and kinematics model of the underactuated unmanned surface vessel is converted into attitude subsystem and position subsystem by redefining the output, which overcomes the problem of insufficient control input of the underactuated unmanned surface vessel; online approximation is performed by using an MLP-based RBF neural network, which effectively solves the problem of parameter uncertainty and external interference in the formation of underactuated unmanned surface vessels; at the same time, the dynamic surface control method is adopted to solve the complex differential explosion problem generated in the calculation process, which significantly reduces the amount of computation and further improves the robustness of the system.

[0143] The effectiveness of this implementation method is verified through data simulation:

[0144] The model ship parameters used in this section are shown in Table 1. Consider a formation of 6 UUSVs, numbered 1-6, and the initial state settings of the formation are shown in Table 2.

[0145] Table 1. Main model parameters of UUSV.

[0146]

[0147] Table 2. Initial state of UUSV formation.

[0148] UUSV parameter Vehicle 1 <![CDATA[[-20m,30m,Orad,0.1m / s,0.1m / s,0.1rad / s] T ]]> Vehicle 2 <![CDATA[[-10m,20m,-π / 4rad,0.1m / s,0.1m / s,0.1rad / s] T ]]> Vehicle 3 <![CDATA[[-15m,5m,Orad,0.11m / s,0.1m / s,0.1rad / s] T ]]> Vehicle 4 <![CDATA[[-25m,10m,Orad,0.1m / s,0.1m / s,0.1rad / s] T ]]> Vehicle 5 <![CDATA[[-10m,5m,Orad,0.1m / s,0.1m / s,0.1rad / s] T ]]> Vehicle 6 <![CDATA[[-5m,5m,0rad,0.1m / s,0.1m / s,0.1rad / s] T ]]>

[0149] The control law and adaptive law designed above were used for simulation verification. Figure 1 The trajectories of the six vehicles on the plane are shown. It is clear that the desired formation shape was eventually obtained, realizing the transformation from the initial state to wedge formation, column formation, and line formation, which is in accordance with Theorem 1 and Theorem 2.

[0150] According to Theorem 1 and Theorem 2, using the designed controller, all signals in the closed-loop system should be SGUUB. Figure 2 The error variable z of the control model for each vehicle is shown. i1 ,z i2 ,z i3 ,z i4 ,z i5 ,z i6 It is evident that the curve exhibits less chattering and is relatively smooth, with the model's error variables all being SGUUB.

[0151] In addition, the control input signal τ of each vehicle X and τ σ like Figure 3 As shown in the figure, τ can be observed in this embodiment. σThe simulation curves are relatively smooth and chatter-free; all control signals for the position and attitude subsystems are continuous and SGUUB. The virtual control laws α1 and α2 for each vehicle are as follows: Figure 4 As shown in the figure, in this embodiment, it can be observed that all control signals of the attitude subsystem and the position subsystem are continuous and SGUUB.

[0152] The simulation results described above all conform to Theorem 1 and Theorem 2.

[0153] The above description of several specific embodiments further details the technical solution provided by the present invention in order to highlight the advantages and benefits of the technical solution provided by the present invention. However, the above-described specific embodiments are not intended to limit the present invention. Any reasonable modifications and improvements to the present invention, combinations of embodiments, and equivalent substitutions based on the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An underactuated USV formation control method based on output redefinition, characterized in that the method Comprising: a step of establishing a kinematic and dynamic model of the underactuated USV, redefining the output of the kinematic and dynamic model of the underactuated USV, and decomposing into an attitude subsystem and a position subsystem; a step of designing an attitude controller and a position controller through the attitude subsystem and the position subsystem respectively, and approximating nonlinear terms of the system; a step of avoiding repeated differentiation of a virtual control law in the model through a first-order filter; a step of combining an adaptive control law and an auxiliary adaptive law based on a minimum learning parameter method; Specifically: introducing a dynamic surface control technology in the control law design, and avoiding repeated differentiation of a virtual control law through a first-order filter; combining a minimum learning parameter method and a radial basis neural network to approximate nonlinear terms of the system, and realizing low-complexity nonlinear compensation; combining an adaptive control law and an auxiliary adaptive law based on a minimum learning parameter method, and realizing parameter dynamic adjustment and disturbance suppression; constructing a Lyapunov function to analyze the stability of the attitude controller and the position controller, and ensuring that the system is globally ultimately uniformly bounded; where, for the attitude subsystem, the error variable is defined as , , as follows wherein is a virtual control law, is a filtered control law, each element of is obtained by a first order filter with time constant is given by the corresponding diagonal element, denotes the yaw rate of the UUSV, is the yaw angle; For the position subsystem, define error variables , , As follows wherein is the position vector of the th UUSV in the earth coordinate system, is the virtual control law, is the filtered control law, each element of is obtained by a first order filter with time constant given by the corresponding diagonal element.

2. The output redefinition based underactuated USV formation control method of claim 1, wherein, further comprising a step of combining a fuzzy control strategy in the attitude controller and the position controller.

3. An underactuated unmanned surface vehicle formation control device based on output redefinition, characterized in that, The device comprises: a module of establishing a kinematic and dynamic model of the underactuated USV, redefining the output of the kinematic and dynamic model of the underactuated USV, and decomposing into an attitude subsystem and a position subsystem; a module of designing an attitude controller and a position controller through the attitude subsystem and the position subsystem respectively, and approximating nonlinear terms of the system; a module of avoiding repeated differentiation of a virtual control law in the model through a first-order filter; a module of combining an adaptive control law and an auxiliary adaptive law based on a minimum learning parameter method; Specifically: introducing a dynamic surface control technology in the control law design, and avoiding repeated differentiation of a virtual control law through a first-order filter; combining a minimum learning parameter method and a radial basis neural network to approximate nonlinear terms of the system, and realizing low-complexity nonlinear compensation; combining an adaptive control law and an auxiliary adaptive law based on a minimum learning parameter method, and realizing parameter dynamic adjustment and disturbance suppression; constructing a Lyapunov function to analyze the stability of the attitude controller and the position controller, and ensuring that the system is globally ultimately uniformly bounded; where, for the attitude subsystem, the error variable is defined as , , as follows wherein is a virtual control law, is a filtered control law, each element of is obtained by a first order filter with time constant is given by the corresponding diagonal element, denotes the yaw rate of the UUSV, is the yaw angle; For the position subsystem, define error variables , , As follows wherein is the position vector of the th UUSV in the geodetic coordinate system, is the virtual control law, is the filtered control law, each element of is obtained by a first order filter with time constant given by the corresponding diagonal element.

4. Computer storage medium for storing a computer program, characterized in that when the computer reads the computer program, the computer executes the method of claim 1.

5. A computer comprising a processor and a storage medium, characterized in that when the processor reads the computer program stored in the storage medium, the computer executes the method of claim 1.

6. Computer program product as computer program, characterized in that when the computer program is executed, the method of claim 1 is realized.

Images