Marine vehicle cluster distributed control method and system based on inverse optimization theory

Abstract

The application discloses a kind of marine vehicle cluster distributed control method and system based on inverse optimization theory, it is related to marine vehicle control technical field to overcome the deficiency of existing state feedback control algorithm.The technical points of the present application include: for the marine vehicle system working in the water surface environment, the multi-marine vehicle system cluster dynamics model with uncertain modeling parameters is constructed;A cluster state tracking error system based on "navigation-following" is constructed, and a global dynamic cost function containing uncertain modeling parameters is defined;A distributed collaborative optimization controller is designed, and the optimal control quantity that makes the cost function optimal is solved using global inverse optimization theory;The optimal control quantity is used to distribute the control of the marine vehicle cluster.The present application can ensure the state consistency of multiple follower marine vehicles and leader marine vehicles and the optimality of the system global cost function under uncertain modeling parameters.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of marine vehicle control, and particularly relates to a marine vehicle cluster distributed control method and system based on inverse optimization theory. BACKGROUND

[0002] The marine robot cluster system has many advantages such as wide working range and strong fault tolerance, and has become one of the indispensable equipment in the field of ocean engineering. Efficient distributed cooperative control technology is the key prerequisite for the marine robot cluster system to complete various work tasks.

[0003] In the prior art, the document "Distributed observer-based formation trajectory tracking method of leader-following multi-AUV system, Ocean Engineering, 2022, 260:11019" discloses a multi-marine vehicle feedback cooperative control method. The method designs a state feedback control algorithm based on an observer, and realizes the "leader-following" state consistency of the multi-marine vehicle. The technical problem existing in the state feedback control algorithm described in the document is that the parameter calculation of the controller needs to rely on the accurate marine vehicle cluster dynamics system model parameters, and the desired control performance cannot be achieved with the optimal control input. SUMMARY

[0004] In view of the above problems, the present application proposes a marine vehicle cluster distributed control method and system based on inverse optimization theory to overcome the shortcomings of the existing state feedback control algorithm which needs to rely on accurate marine vehicle cluster dynamics system model parameters and cannot achieve the desired control performance with the optimal control input.

[0005] According to an aspect of the present application, a marine vehicle cluster distributed control method based on inverse optimization theory is proposed, which comprises:

[0006] Step one, for the marine vehicle system working in the water surface environment, a multi-marine vehicle system cluster dynamics model with uncertain modeling parameters is constructed;

[0007] Step two, for the multi-marine vehicle system cluster dynamics model with uncertain modeling parameters, a "leader-following" cluster state tracking error system is constructed, and a global dynamic cost function containing uncertain modeling parameters is defined;

[0008] Step 3: Design a distributed collaborative optimization controller for the cluster state error system and the global dynamic cost function, and use global inverse optimization theory to solve the optimal control variable that optimizes the cost function;

[0009] Step 4: Use the optimal control quantity to perform distributed control on the ocean vehicle cluster.

[0010] Furthermore, the multi-ocean vehicle system cluster dynamics model with uncertain modeling parameters in step 1 is:

[0011]

[0012] Where x i Represents state variables; Δ1=cos(ψ0)cos(ψ t )-sin(ψ0)sin(ψ t )-cos(ψ0),Δ2=sin(ψ0)cos(ψ t )+cos(ψ0)sin(ψ t )-sin(ψ0) and denote the inertia matrix, fluid damping matrix and mooring force matrix of the ocean vehicle respectively; is the rotation matrix, ψ0 is a time-invariant constant; u i represents the control variable.

[0013] Furthermore, the cluster state tracking error system in step 2 is:

[0014]

[0015] Where N is the total number of aircraft; a ij Represents the communication relationship between the i-th ocean vehicle and the j-th ocean vehicle. When the i-th ocean vehicle can obtain the status information of the j-th ocean vehicle, a ij =1, otherwise a ij =0.

[0016] Furthermore, the global dynamic cost function in step 2 is:

[0017]

[0018] Where, I N represents the N-order identity matrix, R>0 and Q≥0 are symmetric matrices, To control the gain parameter, The full-rank matrix in the Laplacian matrix corresponding to the directed graph describing the communication relationship between multiple ocean vehicles; Represents the global form of state tracking error; Represents the global form of the vehicle cluster control variables.

[0019] Furthermore, the distributed collaborative optimization controller described in step 3 is designed as follows:

[0020]

[0021] Where, is the control gain matrix; the control gain parameters in the optimal control quantity that makes the cost function optimal The following conditions are met:

[0022]

[0023] Where, σ min {Q} represents the minimum singular value of the symmetric matrix Q; Representation matrix The smallest positive singular value of α2 is a constant and satisfies

[0024]

[0025] According to another aspect of the present invention, a distributed control system for a swarm of ocean vehicles based on inverse optimization theory is proposed. The system comprises:

[0026] A dynamic model building module is configured to build a multi-ocean vehicle system cluster dynamic model with uncertain modeling parameters for an ocean vehicle system operating in a water surface environment;

[0027] an error model construction module configured to construct a swarm state tracking error system based on a “leader-follower” strategy for the multi-ocean vehicle system swarm dynamics model with uncertain modeling parameters, and simultaneously define a global dynamic cost function including the uncertain modeling parameters;

[0028] An optimal controller design module is configured to design a distributed collaborative optimization controller for the cluster state error system and the global dynamic cost function, and to use global inverse optimization theory to solve the optimal control variable that optimizes the cost function;

[0029] A distributed control module is configured to perform distributed control on a cluster of ocean vehicles using the optimal control variable.

[0030] Furthermore, the multi-ocean vehicle system cluster dynamics model with uncertain modeling parameters in the dynamics model construction module is:

[0031]

[0032] Where x i Represents state variables; Δ1=cos(ψ0)cos(ψ t )-sin(ψ0)sin(ψ t )-cos(ψ0),Δ2=sin(ψ0)cos(ψ t )+cos(ψ0)sin(ψ t )-sin(ψ0) and denote the inertia matrix, fluid damping matrix and mooring force matrix of the ocean vehicle respectively; is the rotation matrix, ψ0 is a time-invariant constant; u i represents the control variable.

[0033] Furthermore, the cluster state tracking error system in the error model construction module is:

[0034]

[0035] Where N is the total number of aircraft; a ij Represents the communication relationship between the i-th ocean vehicle and the j-th ocean vehicle. When the i-th ocean vehicle can obtain the status information of the j-th ocean vehicle, a ij =1, otherwise a ij =0.

[0036] Furthermore, the global dynamic cost function in the error model construction module is:

[0037]

[0038] Where, I N represents the N-order identity matrix, R>0 and Q≥0 are symmetric matrices, To control the gain parameter, The full-rank matrix in the Laplacian matrix corresponding to the directed graph describing the communication relationship between multiple ocean vehicles; Represents the global form of state tracking error; Represents the global form of the vehicle cluster control variables.

[0039] Furthermore, the distributed collaborative optimization controller in the optimal controller design module is designed as follows:

[0040]

[0041] Where, is the control gain matrix; the control gain parameters in the optimal control quantity that makes the cost function optimal The following conditions must be met:

[0042]

[0043] Where, σ min {Q} represents the minimum singular value of the symmetric matrix Q; Representation matrix The smallest positive singular value of α2 is a constant and satisfies

[0044]

[0045] The beneficial technical effects of the present invention are:

[0046] The present invention proposes a distributed control method and system for a swarm of ocean vehicles based on inverse optimization theory. First, a multi-vessel swarm dynamics model with uncertain modeling parameters is constructed for a swarm of ocean vehicles operating in a surface environment. Then, a leader-follower state tracking error system is constructed for the swarm dynamics model with uncertain modeling parameters, and a global dynamic cost function containing the uncertain modeling parameters is defined. Then, a distributed collaborative optimization controller is designed based on the swarm state error system and the global dynamic cost function. The optimal control variable that optimizes the cost function is solved using global inverse optimization theory. Finally, the swarm of ocean vehicles is distributedly controlled using the optimal control variable. Using the proposed distributed collaborative optimization control method for a swarm of ocean vehicles, each follower ocean vehicle achieves state consistency with the leader ocean vehicle in approximately 10 seconds, and the global dynamic cost function eventually converges to the theoretical optimal value. Therefore, the proposed distributed collaborative optimization controller ensures state consistency between multiple follower ocean vehicles and the leader ocean vehicle under uncertain modeling parameters, as well as the optimality of the system's global cost function. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The present invention can be better understood by referring to the description given below in conjunction with the accompanying drawings, which together with the following detailed description are included in this specification and form a part of this specification, and are used to further illustrate the preferred embodiments of the present invention and explain the principles and advantages of the present invention.

[0048] Figure 1 This is a flow chart of a distributed control method for a cluster of ocean vehicles based on inverse optimization theory according to an embodiment of the present invention;

[0049] Figure 21 is a schematic diagram of a communication topology structure between a leader ocean vehicle (labeled 4) and three follower ocean vehicles (labeled 1 to 3) in an embodiment of the present invention;

[0050] Figure 3 are the four ocean vehicle state components under the control protocol proposed in the embodiment of the present invention || x i ||A graph showing the evolution of the norm over time;

[0051] Figure 4 are the heading angles ψ of the four ocean vehicles under the control protocol proposed in the embodiment of the present invention. i Graphs of the evolution over time;

[0052] Figure 5 It is a graph showing the evolution of the global dynamic cost function over time under the control protocol proposed in the embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to enable those skilled in the art to better understand the present invention, exemplary embodiments or examples of the present invention will be described below with reference to the accompanying drawings. Obviously, the described embodiments or examples are only some of the embodiments or examples of the present invention, and not all of them. Based on the embodiments or examples of the present invention, all other embodiments or examples obtained by those skilled in the art without creative work should fall within the scope of protection of the present invention.

[0054] The present invention provides a distributed collaborative optimization control method and system for a cluster of ocean vehicles. Considering a class of ocean vehicle systems working in a water surface environment, a multi-ocean vehicle system cluster dynamics model with uncertain modeling parameters is constructed; for the multi-ocean robot cluster dynamics model with uncertain modeling parameters, a cluster state tracking error system based on "leader-follower" is constructed, and a global dynamic cost function containing uncertain modeling parameters is defined at the same time; based on the constructed cluster state tracking error system and the global dynamic cost function, a distributed collaborative optimization controller is designed, and the global inverse optimal control theory is used to solve the control gain parameters that can ensure the optimal cost function; the Lyapunov stability theory is used to prove that the designed distributed collaborative optimization controller can ensure the asymptotic stability of the state error tracking system. The present invention can optimize the control input in the presence of uncertain modeling parameters, which is more conducive to the application of the controller in actual systems.

[0055] The embodiment of the present invention proposes a distributed control method for ocean vehicle clusters based on inverse optimization theory, such as Figure 1 As shown, the method includes:

[0056] Step one, for the marine vehicle system working in the water surface environment, a multi-marine vehicle system cluster dynamics model with uncertain modeling parameters is constructed;

[0057] Step two, for the multi-marine vehicle system cluster dynamics model with uncertain modeling parameters, a cluster state tracking error system based on 'leader-following' is constructed, and a global dynamic cost function containing uncertain modeling parameters is defined;

[0058] Step three, for the cluster state error system and the global dynamic cost function, a distributed cooperative optimization controller is designed, and the optimal control quantity making the cost function optimal is solved by using global inverse optimization theory;

[0059] Step four, the optimal control quantity is used for distributed control of the marine vehicle cluster.

[0060] The embodiment of the application is described in detail below.

[0061] Firstly, in step one, considering a kind of marine vehicle system working in the water surface environment, a multi-marine vehicle system cluster dynamics model with uncertain modeling parameters is constructed.

[0062] The following six-degree-of-freedom marine vehicle cluster dynamics model is given:

[0063]

[0064] Among them,

[0065]

[0066] and respectively represent the inertia matrix, the fluid damping matrix and the mooring force matrix of the marine vehicle; r ix and v ix respectively represent the longitudinal position and velocity; r iy and v iy respectively represent the lateral position and velocity; ψ i and ω i respectively represent the heading angle and the heading angle velocity, is a rotation matrix; N represents the total number of vehicles.

[0067] The state variable is selected i Assume that all marine vehicles have the same heading angle, i.e. ψ

[0068]

[0069] In the formula,

[0070]

[0071] Let ψ0 be a time-invariant constant and ψ t =ψ-ψ0, then the nonlinear dynamic model in equation (2) can be rewritten as:

[0072]

[0073] in,

[0074]

[0075] Δ1=cos(ψ0)cos(ψ t )-sin(ψ0)sin(ψ t )-cos(ψ0),Δ2=sin(ψ0)cos(ψ t )+cos(ψ0)sin(ψ t )-sin(ψ0). It is worth noting that since the expressions of Δ1 and Δ2 only involve sine and cosine functions, they are norm-bounded, which means that ΔA is an uncertain modeling parameter with a norm-bounded value. Therefore, Equation (3) can be regarded as a multi-ocean robot cluster dynamics model with uncertain modeling parameters. In addition, the dynamics model of the leader ocean vehicle that only provides reference state trajectory information is defined as x N+1 =(A0+ΔA)x N+1 .

[0076] Then, in step 2, for the linearized dynamic model of the multi-ocean vehicle system cluster with uncertain modeling parameters constructed in step 1, a cluster state tracking error system based on "leader-follower" is constructed, and a global dynamic cost function including uncertain modeling parameters is defined.

[0077] The subscript of the leader vehicle is defined as N+1, and the subscript of the follower vehicle is defined as 1,…,N. The state tracking error system of multiple ocean vehicles based on “leader-follower” is constructed as follows:

[0078]

[0079] Where a ij Represents the communication relationship between the i-th ocean vehicle and the j-th ocean vehicle. When the i-th ocean vehicle can obtain the status information of the j-th ocean vehicle, a ij =1, otherwise a ij =0.

[0080] Using directed graph Describe the communication relationships between multiple ocean vehicles, and A directed graph comprising a directed spanning tree rooted at the leader marine vehicle The corresponding Laplacian matrix can be expressed as follows:

[0081]

[0082] where, is a full rank matrix, and are matrix parameters with appropriate dimensions.

[0083] The global form of equation (4) is expressed using the Kronecker product as:

[0084]

[0085] where, denotes the global form of the state tracking error, I p is an identity matrix with matching dimensions, 1 N denotes an N-dimensional column vector with all elements equal to 1. Taking the derivative of equation (6) gives

[0086]

[0087] where, denotes the global form of the vehicle swarm control variable. Since is a full rank matrix, when the system (7) is stable, there exists , i.e., the multiple follower marine vehicles achieve state agreement with the leader marine vehicle.

[0088] For the system given in equation (7), define the following global dynamic cost function that includes uncertain modeling parameters:

[0089]

[0090] where, R > 0 and Q > 0 are symmetric matrices, is a control gain parameter,

[0091] Then, in step three, for the state error system and the global dynamic cost function constructed in step two, design a distributed cooperative optimization controller, and solve the optimal control quantity that makes the cost function optimal using the global inverse optimization theory, i.e., solve the control gain parameter that can guarantee the optimality of the cost function using the global inverse optimization theory.

[0092] According to equation (7) and equation (8), design the following distributed cooperative optimization controller:

[0093]

[0094] Where, is the control gain matrix.

[0095] According to the optimal control theory, select And the matrix is a positive definite solution to the following algebraic Riccati equation:

[0096]

[0097] Then the distributed collaborative optimization controller (9) can ensure the optimality of the global dynamic cost function (8).

[0098] In addition, when the matrix BR -1 B T When the eigenvalue of is non-zero, ensure that the weight matrix parameters in the global dynamic cost function The conditions for being non-negative are:

[0099]

[0100] Where σ min {Q} represents the smallest singular value of the matrix Q, Representation matrix The minimum singular value of . Considering Therefore, there is Then the conditions for formula (11) to be valid can be expressed as

[0101]

[0102] According to formula (12), when the control gain parameter When the following conditions are met, the weight matrix parameters in the global dynamic cost function are is non-negative:

[0103]

[0104] Where α1 is a constant and satisfies

[0105]

[0106] When the matrix BR -1 B T When the eigenvalue of has zero, the condition in formula (11) will no longer apply, ensuring that the weight matrix parameters in the global dynamic cost function The condition for being non-negative can be restated as

[0107]

[0108] Where, Representation matrix The smallest positive singular value of . Considering that When established Similar to the derivation process of equations (11) to (14), when the control gain parameter When the following conditions are met, the weight matrix parameters in the global dynamic cost function are is non-negative:

[0109]

[0110] Where α2 is a constant and satisfies

[0111]

[0112] Then, in step four, the optimal control quantity is used to perform distributed control on the ocean vehicle cluster.

[0113] Furthermore, Lyapunov stability theory is used to prove that the distributed collaborative optimization controller designed in step three can ensure the asymptotic stability of the cluster state tracking error system in step two as shown below.

[0114] For system (7), the following Lyapunov function is selected:

[0115]

[0116] When the control input When , the derivative of formula (18) is

[0117]

[0118] Will Substituting the expression of into formula (19) we can get

[0119]

[0120] according to And R>0 can be deduced According to Lyapunov stability theory, system (7) is asymptotically stable, that is, Therefore, the distributed collaborative optimization controller in Equation (9) can ensure that multiple follower ocean vehicles can successfully track the state of the leader ocean vehicle.

[0121] The following examples are used to verify the beneficial technical effects of the present invention.

[0122] Assume that there is a leader ocean vehicle and three follower ocean vehicles forming a cluster system, and the communication topology between the ocean vehicles is as follows: Figure 2 As shown, the Laplace matrix corresponding to the communication topology is expressed as follows:

[0123]

[0124] The inertia matrix, fluid damping matrix and mooring force matrix parameters of the selected system are as follows:

[0125]

[0126] Let ψ0=0, R=I3 and Q=1000I6, the parameters of matrixes and can be calculated by using formula (10) as follows:

[0127]

[0128] According to the uncertain modeling parameter ||ΔA||≤0.4405 can be calculated. The control gain parameter can be calculated by using inequality (13) as follows: Therefore, the control gain parameter

[0129] In addition, according to the optimal control theory, the theoretical optimal value of the global dynamic cost function is as follows:

[0130]

[0131] Further, let the initial states of the leader marine vehicle (marked as x4(0)) and the three follower marine vehicles (marked as x i (0), i=1,...,3) be as follows:

[0132]

[0133] Then, the state norm evolution trajectory curve, the heading angle evolution trajectory curve and the global dynamic cost function evolution trajectory curve of the leader marine vehicle and the three follower marine vehicles under the action of the proposed control protocol can be obtained as shown in Figures 3 to 5 The state norm evolution trajectory simulation curve shows that under the action of the proposed distributed cooperative optimization controller, the tracking error norm of each state component gradually converges to zero at about 10 seconds. The heading angle evolution trajectory simulation curve shows that the four marine vehicles all have the same heading angle, which is consistent with the precondition in step one. The global dynamic cost function evolution trajectory curve shows that under the action of the proposed distributed cooperative optimization controller, the global dynamic cost function finally converges to 5951.1, which is consistent with the theoretical optimal value. Therefore, the control method proposed in the present application can realize distributed cooperative optimization control of a multi-marine vehicle cluster system in the presence of uncertain modeling parameters.

[0134] It should be noted that the contents (such as graph theory, optimal control theory, Lyapunov stability theory, matrix theory) not described in detail in the embodiments of the present application belong to the common common sense in the field.

[0135] Another embodiment of the present application provides a marine vehicle cluster distributed control system based on inverse optimization theory, which comprises:

[0136] a dynamics model construction module configured to construct a multi-marine vehicle system cluster dynamics model with uncertain modeling parameters for marine vehicle systems working in a water surface environment;

[0137] an error model construction module configured to construct a "leader-follower" based cluster state tracking error system for the multi-marine vehicle system cluster dynamics model with uncertain modeling parameters, and define a global dynamic cost function containing uncertain modeling parameters;

[0138] an optimal controller design module configured to design a distributed collaborative optimization controller for the cluster state error system and the global dynamic cost function, and solve the optimal control quantity that makes the cost function optimal by using global inverse optimization theory;

[0139] a distributed control module configured to perform distributed control on the marine vehicle cluster by using the optimal control quantity.

[0140] In the embodiment, preferably, the multi-marine vehicle system cluster dynamics model with uncertain modeling parameters in the dynamics model construction module is:

[0141]

[0142] wherein,

[0143]

[0144] Δ1=cos(ψ0)cos(ψ t )-sin(ψ0)sin(ψ t )-cos(ψ0),Δ2=sin(ψ0)cos(ψ t )+cos(ψ0)sin(ψ t )-sin(ψ0) and respectively represent the inertia matrix, the fluid damping matrix and the mooring force matrix of the marine vehicle; is a rotation matrix, ψ0 is a time-invariant constant; x i represents a state variable, u i represents a control variable.

[0145] In this embodiment, preferably, the cluster state tracking error system in the error model building module is:

[0146]

[0147] Where N is the total number of aircraft; a ij Represents the communication relationship between the i-th ocean vehicle and the j-th ocean vehicle. When the i-th ocean vehicle can obtain the status information of the j-th ocean vehicle, a ij =1, otherwise a ij =0.

[0148] In this embodiment, preferably, the global dynamic cost function in the error model construction module is:

[0149]

[0150] Where, I N represents the N-order identity matrix, R>0 and Q≥0 are symmetric matrices, To control the gain parameter, The full-rank matrix in the Laplacian matrix corresponding to the directed graph describing the communication relationship between multiple ocean vehicles; Represents the global form of state tracking error; Represents the global form of the vehicle cluster control variables.

[0151] In this embodiment, preferably, the distributed collaborative optimization controller in the optimal controller design module is designed as follows:

[0152]

[0153] Where, is the control gain matrix; the control gain parameters in the optimal control quantity that makes the cost function optimal The following conditions are met:

[0154]

[0155] Where, σ min {Q} represents the minimum singular value of the symmetric matrix Q; Representation matrix The smallest positive singular value of α2 is a constant and satisfies

[0156]

[0157] The functions of the ocean vehicle cluster distributed control system based on inverse optimization theory described in an embodiment of the present invention can be explained by the aforementioned ocean vehicle cluster distributed control method based on inverse optimization theory. Therefore, for the parts not described in detail in the system embodiment, please refer to the above method embodiment and will not be repeated here.

[0158] Although the present invention has been described with respect to a limited number of embodiments, those skilled in the art, having benefit of the foregoing description, will appreciate that other embodiments are contemplated within the scope of the invention thus described. This disclosure is intended to be illustrative rather than restrictive of the scope of the invention, which is defined by the appended claims.

Claims

1. A distributed control method for ocean vehicle clusters based on inverse optimization theory, characterized in that: include: Step 1: For the ocean vehicle system working in the water surface environment, a multi-ocean vehicle system cluster dynamics model with uncertain modeling parameters is constructed; Step 2: For the multi-ocean vehicle system swarm dynamics model with uncertain modeling parameters, a swarm state tracking error system based on "leader-follower" is constructed, and a global dynamic cost function including the uncertain modeling parameters is defined; Step 3: Design a distributed collaborative optimization controller for the cluster state error system and the global dynamic cost function, and use global inverse optimization theory to solve the optimal control variable that optimizes the cost function; Step 4: Use the optimal control quantity to perform distributed control on the ocean vehicle cluster.

2. The method for distributed control of ocean vehicle clusters based on inverse optimization theory according to claim 1, characterized in that: The multi-ocean vehicle system cluster dynamics model with uncertain modeling parameters described in step 1 is: Where x i Represents state variables; Δ1=cos(ψ0)cos(ψ t )-sin(ψ0)sin(ψ t )-cos(ψ0),Δ2=sin(ψ0)cos(ψ t )+cos(ψ0)sin(ψ t )-sin(ψ0) and denote the inertia matrix, fluid damping matrix and mooring force matrix of the ocean vehicle respectively; is the rotation matrix, ψ0 is a time-invariant constant; u i represents the control variable.

3. The method for distributed control of ocean vehicle clusters based on inverse optimization theory according to claim 2, characterized in that: The cluster state tracking error system described in step 2 is: Where N is the total number of aircraft; a ij Represents the communication relationship between the i-th ocean vehicle and the j-th ocean vehicle. When the i-th ocean vehicle can obtain the status information of the j-th ocean vehicle, a ij =1, otherwise a ij =0.

4. The method for distributed control of ocean vehicle clusters based on inverse optimization theory according to claim 3, characterized in that: The global dynamic cost function in step 2 is: Where, I N represents the N-order identity matrix, R>0 and Q≥0 are symmetric matrices, To control the gain parameter, The full-rank matrix in the Laplacian matrix corresponding to the directed graph describing the communication relationship between multiple ocean vehicles; Represents the global form of state tracking error; Represents the global form of the vehicle cluster control variables.

5. The method for distributed control of ocean vehicle clusters based on inverse optimization theory according to claim 4, characterized in that: The distributed collaborative optimization controller described in step 3 is designed as follows: Where, is the control gain matrix; the control gain parameters in the optimal control quantity that makes the cost function optimal The following conditions must be met: Where, σ min {Q} represents the minimum singular value of the symmetric matrix Q; Representation matrix The smallest positive singular value of α2 is a constant and satisfies 6. A distributed control system for ocean vehicle swarms based on inverse optimization theory, characterized in that: include: A dynamic model building module is configured to build a multi-ocean vehicle system cluster dynamic model with uncertain modeling parameters for an ocean vehicle system operating in a water surface environment; an error model construction module configured to construct a swarm state tracking error system based on a "leader-follower" strategy for the multi-ocean vehicle system swarm dynamics model with uncertain modeling parameters, and simultaneously define a global dynamic cost function including the uncertain modeling parameters; An optimal controller design module is configured to design a distributed collaborative optimization controller for the cluster state error system and the global dynamic cost function, and to use global inverse optimization theory to solve the optimal control variable that optimizes the cost function; A distributed control module is configured to perform distributed control on a cluster of ocean vehicles using the optimal control variable.

7. The ocean vehicle swarm distributed control system based on inverse optimization theory according to claim 6, characterized in that: The multi-ocean vehicle system cluster dynamics model with uncertain modeling parameters in the dynamics model building module is: Where x i Represents state variables; Δ1=cos(ψ0)cos(ψ t )-sin(ψ0)sin(ψ t )-cos(ψ0),Δ2=sin(ψ0)cos(ψ t )+cos(ψ0)sin(ψ t )-sin(ψ0) and denote the inertia matrix, fluid damping matrix and mooring force matrix of the ocean vehicle respectively; is the rotation matrix, ψ0 is a time-invariant constant; u i represents the control variable.

8. The ocean vehicle swarm distributed control system based on inverse optimization theory according to claim 7, characterized in that: The cluster state tracking error system in the error model construction module is: Where N is the total number of aircraft; a ij Represents the communication relationship between the i-th ocean vehicle and the j-th ocean vehicle. When the i-th ocean vehicle can obtain the status information of the j-th ocean vehicle, a ij =1, otherwise a ij =0.

9. The ocean vehicle swarm distributed control system based on inverse optimization theory according to claim 8, characterized in that: The global dynamic cost function in the error model construction module is: Where, I N represents the N-order identity matrix, R>0 and Q≥0 are symmetric matrices, To control the gain parameter, The full-rank matrix in the Laplacian matrix corresponding to the directed graph describing the communication relationship between multiple ocean vehicles; Represents the global form of state tracking error; Represents the global form of the vehicle cluster control variables.

10. The ocean vehicle cluster distributed control system based on inverse optimization theory according to claim 9, characterized in that: The distributed collaborative optimization controller in the optimal controller design module is designed as follows: Where, is the control gain matrix; the control gain parameters in the optimal control quantity that makes the cost function optimal The following conditions must be met: Where, σ min {Q} represents the minimum singular value of the symmetric matrix Q; Representation matrix The smallest positive singular value of α2 is a constant and satisfies

Images