Ship anti-spoofing attack control method based on self-feedback expansion state observer

Abstract

The invention discloses a ship anti-spoofing attack control method based on a self-feedback expansion state observer. The ship anti-spoofing attack control method comprises the steps that a guidance virtual ship and a dynamic virtual ship are set to construct an L2-DVS guidance strategy; constructing a spoofing attack additional signal model of the unmanned ship, and obtaining a ship position signal under the spoofing attack according to the kinematic model of the target unmanned ship; acquiring a longitudinal velocity virtual control law of the target unmanned ship and a spoofing attack compensation law of a yaw angle according to the constructed error model so as to obtain a yaw velocity virtual control law / adaptive law; and constructing a state memory type event triggering mechanism, obtaining a controller / adaptive law of a rudder angle / rotating speed of the target unmanned ship in combination with a self-feedback compensation reduced order expansion state observer, and realizing anti-spoofing attack control of the target unmanned ship according to the yaw velocity virtual control law / adaptive law and the controller / adaptive law of the rudder angle / rotating speed. The problems that the research of the application in the current network attack environment is not complete enough, the unmanned ship is disturbed all the time in the marine environment, the actuator continuously responds to the disturbance, and the excessive wear of the steering engine is greatly increased are solved.

Description

Technical Field

[0001] The present invention relates to the field of unmanned autonomous ship control network security and control engineering, and in particular to a ship anti-deception attack control method based on a self-feedback extended state observer. Background Art

[0002] In the past, in the field of unmanned ship motion control, the unmanned ship system mainly consisted of two parts: reference signal acquisition and controller design. After selecting waypoints to plan the route, the guidance part generates a real-time reference signal through a virtual guide ship. The controller design uses the backstepping method to design a suitable control law based on the error model between the real ship signal and the reference signal to make it converge. [1,2,3] .

[0003] In addition, the extended state observer has been widely used in active disturbance rejection control and compensation of model unknown terms. [4] , and has been used for interference observation in the field of unmanned ships. Its advantage is that the nonlinear part can be estimated together with the random interference and compensated back to the controller to make the controller [5-7] ,The design process is simpler, but due to the inherent characteristics of the observer, it has high requirements for the accuracy of the system state, and the research on its application in the network attack environment is not yet complete. At the same time, since the unmanned ship will always be disturbed in the marine environment and the actuator will continue to respond to these disturbances, especially the steering gear will cause excessive wear. However, the existing unmanned ship control algorithm still has the following shortcomings:

[0004] 1) In the estimation of nonlinear and random interference, the extended state observer in the past relied on the feedback of the first-order estimation error. However, in the case of a deception attack, the first-order position signal is no longer credible, which has imposed certain limitations on the application of the extended state observer in network attack environments.

[0005] 2) Due to the mutual coupling of the kinematic model signals, the observed state is estimated in the form of a matrix. However, in the actual control process, due to the under-actuated characteristics of the unmanned ship, there is no control input in the y direction, which results in a large waste of system computing resources and greatly increases unnecessary design difficulty.

[0006] 3) The traditional mixed threshold event trigger mechanism only considers the state change at the previous moment. When the control input oscillates slightly, the event may be triggered frequently. When the control command curve is at its peak, the signal may be released in small quantities. [8] . Summary of the invention

[0007] The present invention provides a ship anti-deception attack control method based on a self-feedback extended state observer to overcome the above technical problems.

[0008] In order to achieve the above object, the technical solution of the present invention is:

[0009] A ship anti-deception attack control method based on a self-feedback extended state observer specifically comprises the following steps:

[0010] S1: Setting the guided virtual ship and dynamic virtual ship to construct the L2-DVS guidance strategy;

[0011] S2: Obtain the kinematic model and nonlinear dynamic model of the target unmanned ship;

[0012] S3: Construct an additional signal model for the deception attack of the unmanned ship, and obtain the ship position signal under the deception attack according to the kinematic model of the target unmanned ship;

[0013] S4: Based on the kinematic model, according to the ship position signal and L2-DVS guidance strategy, the error model of the target unmanned ship in the attached coordinate system is obtained;

[0014] According to the error model, the virtual control law of the longitudinal speed of the target unmanned ship and the deception attack compensation law of the yaw angle are constructed;

[0015] S5: By introducing the dynamic surface control technology, the yaw rate virtual control law / adaptive law is obtained according to the longitudinal velocity virtual control law and the deception attack compensation law;

[0016] S6: Based on the nonlinear dynamic model of the target unmanned ship, a reduced-order extended state observer with self-feedback compensation is constructed;

[0017] S7: Construct a state memory type event trigger mechanism, and combine it with the reduced-order extended state observer with self-feedback compensation to obtain the controller / adaptive law of the rudder angle / speed of the target unmanned ship;

[0018] S8: According to the yaw angular velocity virtual control law / adaptive law, the rudder angle / rotation speed controller / adaptive law, the anti-deception attack control of the target unmanned ship is realized.

[0019] Furthermore, the L2-DVS guidance strategy constructed in S1 is specifically:

[0020] Setting a guided virtual ship for obtaining a reference path of a target unmanned ship and a dynamic virtual ship for guiding the movement of the target unmanned ship;

[0021] The expression of the guided virtual ship model is:

[0022]

[0023] The expression of the dynamic virtual ship model is:

[0024]

[0025] Where: x g ,y g ,ψ g They respectively represent the longitudinal position, lateral position and heading angle of the guided virtual ship; Respectively represent x g ,y g ,ψ g The first derivative of u g ,v g Respectively represent the longitudinal speed and lateral speed of the guided virtual ship; r d represents the yaw angular velocity of the guided virtual ship, that is, the reference yaw angular velocity of the target unmanned ship; x d ,y d They represent the longitudinal position and lateral position of the dynamic virtual ship, i.e., the position reference signal of the target unmanned ship; ψ rd ,u d ,v d They represent the yaw angle, longitudinal velocity and lateral velocity of the dynamic virtual ship respectively; and Represents x d ,y d The first derivative of d represents the yaw angle of the dynamic virtual ship, i.e., the yaw direction reference signal of the target unmanned ship; x, y represent the longitudinal position and lateral position of the target unmanned ship in the geodetic coordinate system respectively;

[0026] Get the relative distance z between the target unmanned ship and the guided virtual ship e ;

[0027] The relative distance z e The expression is

[0028]

[0029] Set the distance threshold L d , compare the relative distance z e With distance threshold L d size;

[0030] If the relative distance z is confirmed e Less than the distance threshold L d That is z e v d , then the position signal of the guided virtual ship is used as the position reference signal of the target unmanned ship, that is,

[0031] x d =x g ,y d =y g

[0032] If the relative distance z is confirmed e Greater than or equal to the distance threshold L d That is z e ≥L d , then according to the distance threshold L between the guided virtual ship and d , obtain the position reference signal of the target unmanned ship, that is

[0033] x d =x g +L d cosψ d ,y d =y g +sinψ d .

[0034] Furthermore, the kinematic model and nonlinear dynamic model of the target unmanned ship obtained in S2; the expression of the kinematic model is:

[0035]

[0036] The expression of the nonlinear dynamic model is

[0037]

[0038] Where: x, y, ψ represent the ship position and heading signal of the target unmanned ship respectively; u, v, r represent the longitudinal velocity, lateral velocity and yaw angular velocity of the target unmanned ship respectively; n, δ represent the control input provided by the host and the steering gear respectively; T u (·) and F r (·) represents the control gain function; d wu ,d wv ,d wr Respectively represent the interference items caused by wind, waves and currents in the ocean environment; m u ,m v ,m r ,d u1 ,d v1 ,d r1 ,d u2 ,d v2 ,d r2 ,d u3 ,d v3 ,d r3 represents the unknown parameters of the unmanned ship's inertia, hydrodynamic damping, and nonlinear damping terms; f u (υ),f v (υ),f r (υ) represents the nonlinear function of higher-order hydrodynamic effects.

[0039] Furthermore, the S3 specifically includes the following steps:

[0040] S31: Construct an additional signal model for the deception attack of unmanned ships, which is expressed as follows:

[0041]

[0042] Where: i represents the additional deception attack signal in the x,y direction of the target unmanned ship, and i = [x, y] T , a i ,b i , represents a non-zero constant of the design and a i =[a x ,a y ,a ψ ] T ,b i =[a x ,a y ,a ψ ] T , Indicates the use of trigonometric functions to simulate nonlinear functions with unknown parameters;

[0043] S32: According to the kinematic model of the target unmanned ship combined with the deception attack signal of the deception attack additional signal model, the ship position signal under the deception attack is obtained, and its expression is:

[0044]

[0045] Furthermore, the S4 specifically includes the following steps:

[0046] S41: According to the ship position signal and the L2-DVS guidance strategy, the error model of the target unmanned ship in the attached coordinate system is obtained;

[0047] The expression of the error model is:

[0048]

[0049] Where: x ,x y Denote the deception attack signal Ξ x With Ξ y The first derivative of x e ,y e ,ψ e Indicates the longitudinal position error, lateral position error and heading angle error of the target unmanned ship;

[0050] S42: Construct the first Lyapunov function V according to the error model 1 , to obtain the derivative of the first Lyapunov function

[0051] The first Lyapunov function V 1 for

[0052]

[0053] The derivative of the first Lyapunov function is for

[0054]

[0055] S43: In order to satisfy the derivative of the first Lyapunov function Stability, construct the virtual control law of the longitudinal speed of the target unmanned ship and the deception attack compensation law of the yaw angle, and its expression is

[0056]

[0057] Where: Respectively represent χ x ,x y The estimated value of u represents the virtual control law of longitudinal velocity; K represents the deception attack compensation law of yaw angle; u ,K r1 Represents design parameters.

[0058] Furthermore, the S5 specifically includes the following steps:

[0059] S51: By introducing the dynamic surface control technology, the longitudinal speed virtual control law and the deception attack compensation law are filtered and the expression is:

[0060]

[0061] α i (0) = β i (0)

[0062] i=r,u,ψ e

[0063] Where: β i represents a first-order filter; α i It is a general term for the virtual control law of longitudinal velocity, the deception attack compensation law and the virtual control law of yaw angular velocity; τ i represents the time constant of the first-order filter; α i (0),β i (0) indicates α i With β i The initial value of

[0064] S52: Define the first intermediate parameter y according to step S51 u , And y u =α u -β u ,

[0065] And construct the second Lyapunov function V based on the intermediate parameter 2 , whose expression is

[0066]

[0067] Where: represents the adjustable positive design parameter of the second Lyapunov function; Represents χ x ,x y The estimation error of

[0068] Get the second Lyapunov function V 2 The derivative of Its expression is

[0069]

[0070] Where: τ u , represents the time constant of a first-order filter;

[0071] S53: In order to satisfy the derivative of the second Lyapunov function Stability, construct the virtual control law / adaptive law of the yaw angular velocity of the target unmanned ship, and its expression is

[0072]

[0073] Where: α r represents the virtual control law of yaw rate; K r represents the design parameters; express The first derivative of r It means that the virtual control law α of the yaw angular velocity is controlled by the dynamic surface control technology. r Time constant for first-order filtering.

[0074] Furthermore, the reduced-order extended state observer for self-feedback compensation constructed in S6 is expressed as

[0075]

[0076] Where: G = [T u ,F r ] T,i=[u,r] T ,U=[|n|n,δ] T ; L 1 ,L 2 represents the adjustable parameter of the design, J i represents the extended state of the extended state observer and E i represents the velocity observation error and represents the expansion state observation error and Indicates J i An estimated value of represents the estimated value of i.

[0077] Furthermore, the S7 specifically includes the following steps:

[0078] S71: The constructed state memory event trigger mechanism includes:

[0079] S711: Define the state change E of the target unmanned ship's rudder angle / speed u , whose expression is

[0080] E u =η 1 E pu +η 2 E qu

[0081] E pu =U k (t)-U(t)

[0082] E qu =U k (t)-U(t k -τ)

[0083] Where: η 1 ,η 2 represents a positive real number and η 1 +η 2 =1, ξ 1 ,ξ 2 represents the adjustable parameters of the design; E pu ,E qu Indicates the intermediate parameter; U k (t),U(t),U(t k -τ) represent the rudder angle / speed of the target unmanned ship when it is triggered at time k, the continuous output signal of the rudder angle / speed of the target unmanned ship, and the continuous output signal of the rudder angle / speed of the target unmanned ship at time t k -Rudder angle / speed when triggered at time τ;

[0084] S712: According to the state change amount Eu Design a state memory type event trigger mechanism, whose expression is t k+1 = inf{t>t k ||E u (t)|≥ξ 3 |U(t)|+ξ 4}

[0085] Where: t k+1 ,t k represents the time parameter; 3 ,ξ 4 represents the adjustable parameters of the design;

[0086] S713: Define the second intermediate parameter Ω i and

[0087] And according to the second intermediate parameter combined with the state memory type event trigger mechanism, the control input of the target unmanned ship is obtained, and its expression is:

[0088]

[0089] Where: 1 ,λ 2 represents the adjustable parameters of the design; Represents Ω i The estimated value of α u represents the virtual control law of the longitudinal velocity of the target unmanned ship; i represents the intermediate parameter and

[0090] S72: Based on the dynamic model of the target unmanned ship and step S51, define the second intermediate parameter u e ,r e andu e =β u -u,r e =β r -r;

[0091] And according to the second intermediate parameter and the control input of the target unmanned ship, the third Lyapunov function V is constructed 3 , whose expression is

[0092]

[0093] Where: represents the design parameters of the third Lyapunov function; Represents the second intermediate parameter Ω u ,Ω r The estimation error of Represents χ u,x r The estimation error of

[0094] Get the third Lyapunov function V 3 The derivative of Its expression is

[0095]

[0096] S73: In order to satisfy the derivative of the third Lyapunov function Stability, based on the reduced-order extended state observer with self-feedback compensation, the controller / adaptive law of the rudder angle / speed of the target unmanned ship is constructed, and its expression is

[0097]

[0098] Where: α n Indicates the speed of the target unmanned ship controller; α δ represents the rudder angle of the target unmanned ship controller; k u2 ,k r3 represents an adjustable design parameter; represents the estimated value of u,r by the reduced-order extended state observer compensated by self-feedback; Represents χ u ,x r An estimated value of represents the design parameters; Respectively represent Ω r ,Ω u Estimated value of Ω u (0),Ω y (0) respectively represent Ω r ,Ω u The initial value of represents the design parameter; y (0),x u (0) respectively represent χ y ,x u The initial value of express The first derivative of express The first derivative of .

[0099] Beneficial effects: The present invention provides a ship anti-cheating attack control method based on a self-feedback extended state observer, and the beneficial effects are as follows:

[0100] 1) The present invention aims at online compensation of deception attack signals. The proposed self-feedback extended state observer abandons the correction of the erroneous second-order position signal error, fundamentally avoiding the influence of deception attacks on error estimation, and greatly improving the observation capability of the extended state of the unmanned ship.

[0101] 2) The influence of the deception attack signal is decomposed into the coordinate system of the unmanned ship's appendage, and the yaw angular velocity virtual control law / adaptive law is obtained according to the constructed longitudinal velocity virtual control law and the deception attack compensation law. A set of adaptive parameters is designed for the rate of change of the target unmanned ship with respect to time, which can accurately compensate the control of the target unmanned ship under the deception attack.

[0102] 3) The L1 guidance principle of drones is improved and applied to the guidance of under-actuated unmanned ships, and the L2-DVS guidance strategy is constructed in combination with the DVS guidance principle. At the route generation level, it can avoid the problem of accumulation of calculation errors of the bow angle reference velocity signal over time, and produce a smoother reference path. At the level of guiding unmanned ships, when the distance between the real ship and the logical virtual ship is too large, the target unmanned ship can be guided by the virtual ship to approach the logical virtual ship in a smooth route.

[0103] 4) The state memory event trigger mechanism proposed in the present invention compensates for the problem that the traditional mixed threshold event trigger mechanism only considers the state of the previous moment and releases a small amount of control signal due to the reduced change rate before and after the peak by increasing the use of past control signal state values, and effectively reduces excessive wear of the actuator. BRIEF DESCRIPTION OF THE DRAWINGS

[0104] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces 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 creative labor.

[0105] Figure 1 It is a flow chart of the ship anti-cheating attack control method based on the self-feedback extended state observer of the present invention;

[0106] Figure 2 This is a system block diagram of ship anti-deception attack control in this embodiment;

[0107] Figure 3 Schematic diagram of the principle of the L2-DVS guidance strategy in this embodiment;

[0108] Figure 4 This is the trajectory diagram of the unmanned ship path tracking under the strong deception attack in this embodiment;

[0109] Figure 5 is a curve diagram of the system position heading position error variation in this embodiment;

[0110] Figure 6 This is a speed observation effect diagram in this embodiment;

[0111] Figure 7 This is a diagram of the nonlinearity and disturbance observation effect of the unmanned ship model in this embodiment;

[0112] Figure 8 This is a deception attack adaptive compensation response diagram in this embodiment;

[0113] Fig. 9 This is an event triggering effect diagram of the state memory type event triggering mechanism in this embodiment. DETAILED DESCRIPTION

[0114] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0115] This embodiment provides a ship anti-cheating attack control method based on a self-feedback extended state observer, such as Figure 1 to Figure 2 As shown, the specific steps include:

[0116] S1: Setting the guided virtual ship and dynamic virtual ship to construct the L2-DVS guidance strategy;

[0117] Specifically, in the engineering context of the target unmanned ship USV path tracking control, the reference path is generally generated by connecting waypoints, but the USV cannot complete fixed-point steering. Therefore, the improved L1 guidance is introduced. [9] The L2-DVS guidance strategy composed of the DVS guidance enables the USV to navigate smoothly; and the basic principle of the L2-DVS guidance strategy is as follows Figure 3 As shown, it can be seen that the reference signal is generated by two virtual ships, namely the guided virtual ship (GVS) and the dynamic virtual ship (DVS), and the GVS is used to generate a smooth reference path for the target unmanned ship USV, and the DVS is used to guide the USV heading;

[0118] In a specific embodiment, the constructed L2-DVS guidance strategy is specifically as follows:

[0119] Setting a guided virtual ship for obtaining a reference path of a target unmanned ship and a dynamic virtual ship for guiding the movement of the target unmanned ship;

[0120] The expression of the guided virtual ship model is:

[0121]

[0122] The expression of the dynamic virtual ship model is:

[0123]

[0124] Where: x g ,y g ,ψ g They respectively represent the longitudinal position, lateral position and heading angle of the guided virtual ship; Respectively represent x g ,y g ,ψ g The first derivative of u g ,v g Respectively represent the longitudinal speed and lateral speed of the guided virtual ship; r d represents the yaw angular velocity of the guided virtual ship, that is, the reference yaw angular velocity of the target unmanned ship; x d ,y d They represent the longitudinal position and lateral position of the dynamic virtual ship, i.e., the position reference signal of the target unmanned ship; ψ rd ,u d ,v d They represent the yaw angle, longitudinal velocity and lateral velocity of the dynamic virtual ship respectively; and Represents x d ,y d The first derivative of d represents the yaw angle of the dynamic virtual ship, i.e., the yaw direction reference signal of the target unmanned ship; x, y represent the longitudinal position and lateral position of the target unmanned ship in the geodetic coordinate system respectively;

[0125] Get the relative distance z between the target unmanned ship and the guided virtual ship e , that is, let z e Indicates the actual distance between USV and GVS;

[0126] The relative distance z e The expression is

[0127]

[0128] Set the distance threshold L d , compare the relative distance z e With distance threshold L d size;

[0129] If the relative distance z is confirmed e Less than the distance threshold L d That is z e <L d , then the position signal of the guided virtual ship is used as the position reference signal of the target unmanned ship, that is,

[0130] xd =x g ,y d =y g

[0131] If the relative distance z is confirmed e Greater than or equal to the distance threshold L d That is z e ≥L d , then according to the distance threshold L between the guided virtual ship and d , obtain the position reference signal of the target unmanned ship, that is

[0132] x d =x g +L d cosψ d ,y d =y g +sinψ d ;

[0133] S2: Obtain the kinematic model and nonlinear dynamic model of the target unmanned ship;

[0134] The expression of the kinematic model is

[0135]

[0136] The expression of the nonlinear dynamic model is

[0137]

[0138] Where: x, y, ψ represent the ship position and heading signal of the target unmanned ship respectively; u, v, r represent the longitudinal velocity, lateral velocity and yaw angular velocity of the target unmanned ship respectively; n, δ represent the control input provided by the host and the steering gear respectively; T u (·) and F r (·) represents the control gain function; d wu ,d wv ,d wr Respectively represent the interference items caused by wind, waves and currents in the ocean environment; m u ,m v ,m r ,d u1 ,d v1 ,d r1 ,d u2 ,d v2 ,d r2 ,d u3 ,d v3 ,d r3 represents the unknown parameters of the unmanned ship's inertia, hydrodynamic damping, and nonlinear damping terms; f u (υ),f v(υ),f r (υ) represents the nonlinear function of higher-order hydrodynamic effects;

[0139] S3: Construct an additional signal model for the deception attack of the unmanned ship, and obtain the ship position signal under the deception attack according to the kinematic model of the target unmanned ship;

[0140] The specific steps include:

[0141] S31: Construct an additional signal model for the deception attack of unmanned ships, which is expressed as follows:

[0142]

[0143] Where: i represents the additional deception attack signal in the x,y direction of the target unmanned ship, and i = [x, y] T , a i ,b i , represents a non-zero constant of the design and a i =[a x ,a y ,a ψ ] T ,b i =[a x ,a y ,a ψ ] T , Indicates the use of trigonometric functions to simulate nonlinear functions with unknown parameters;

[0144] S32: This embodiment sends the ship position signal to the ship station based on the geodetic coordinate system satellite, and is subject to information injection type deception attack in the process

[10] ,Thus, we can get the actual received position signal, which is the deception attack signal based on the kinematic model of the target unmanned ship combined with the deception attack additional signal model, and obtain the ship position signal under the deception attack, and its expression is:

[0145] x a =x+Ξ x

[0146] y a =y+Ξ y

[0147] S4: Based on the kinematic model, according to the ship position signal and L2-DVS guidance strategy, the error model of the target unmanned ship in the attached coordinate system is obtained; according to the error model, the virtual control law of the longitudinal speed of the target unmanned ship and the deception attack compensation law of the yaw angle are constructed;

[0148] The specific steps include:

[0149] S41: According to the ship position signal and the L2-DVS guidance strategy, the error model of the target unmanned ship in the attached coordinate system is obtained;

[0150] The expression of the error model is:

[0151]

[0152] Where: x ,x y Denote the deception attack signal Ξ x With Ξ y The first derivative of x e ,y e ,ψ e Indicates the longitudinal position error, lateral position error and heading angle error of the target unmanned ship;

[0153] S42: Construct the first Lyapunov function V according to the error model 1 , to obtain the derivative of the first Lyapunov function

[0154] The first Lyapunov function V 1 for

[0155]

[0156] The derivative of the first Lyapunov function is for

[0157]

[0158] S43: In order to satisfy the derivative of the first Lyapunov function Stability, construct the virtual control law of the longitudinal speed of the target unmanned ship and the deception attack compensation law of the yaw angle, and its expression is

[0159]

[0160] Where: Respectively represent X x ,X y The estimated value of u represents the virtual control law of longitudinal velocity; K represents the deception attack compensation law of yaw angle; u ,K r1 represents the design parameters, and

[0161]

[0162] in: represents an adjustable parameter; Xx (0),X y (0) respectively represent X x ,x y The initial state value of

[0163] S5: By introducing the dynamic surface control technology, the yaw rate virtual control law / adaptive law is obtained according to the longitudinal velocity virtual control law and the deception attack compensation law;

[0164] The specific steps include:

[0165] S51: In order to avoid the exponential explosion problem caused by multiple derivations in the subsequent derivation process, the dynamic surface control technology is introduced. In this embodiment, two first-order filters β are used. u and To replace α u and The longitudinal velocity virtual control law and the deception attack compensation law are filtered and the expression is:

[0166]

[0167] α i (0) = β i (0)

[0168] i=r,u,ψ e

[0169] Where: β i represents a first-order filter; α i It is a general term for the virtual control law of longitudinal velocity, the deception attack compensation law and the virtual control law of yaw angular velocity; τ i represents the time constant of the first-order filter; α i (0),β i (0) indicates α i With β i The initial value of

[0170] S52: Define the first intermediate parameter y according to step S51 u , And y u =α u -β u , and

[0171] And construct the second Lyapunov function V based on the intermediate parameter 2 , whose expression is

[0172]

[0173] Where: represents the adjustable positive design parameter of the second Lyapunov function; Represents χ x ,x y The estimation error of

[0174] Get the second Lyapunov function V 2 The derivative of Its expression is

[0175]

[0176] Where: τ u , represents the time constant of a first-order filter;

[0177] S53: In order to satisfy the derivative of the second Lyapunov function Stability, construct the virtual control law / adaptive law of the yaw angular velocity of the target unmanned ship, and its expression is

[0178]

[0179] Where: α r represents the virtual control law of yaw rate; K r represents the design parameters; express The first derivative of r It means that the virtual control law α of the yaw angular velocity is controlled by the dynamic surface control technology. r The time constant for first-order filtering;

[0180] S6: According to the nonlinear dynamic model of the target unmanned ship, a reduced-order extended state observer with self-feedback compensation is constructed, and the expression of the reduced-order extended state observer with self-feedback compensation is:

[0181]

[0182] Where: G = [T u ,F r ] T ,i=[u,r] T ,U=[|n|n,δ] T ; L 1 ,L 2 represents the adjustable parameter of the design, J i represents the extended state of the extended state observer and E i represents the velocity observation error and represents the expansion state observation error and Indicates J iAn estimated value of represents the estimated value of i;

[0183] S7: Construct a state memory type event trigger mechanism, and combine it with the reduced-order extended state observer with self-feedback compensation to obtain the controller / adaptive law of the rudder angle / speed of the target unmanned ship;

[0184] S71: This embodiment introduces a state memory type event trigger mechanism to solve the problem of excessive wear of the actuator, specifically including:

[0185] S711: Define the state change E of the target unmanned ship's rudder angle / speed u , whose expression is

[0186] E u =η 1 E pu +η 2 E qu

[0187] E pu =U k (t)-U(t)

[0188] E qu =U k (t)-U(t k -τ)

[0189] Where: η 1 ,η 2 represents a positive real number and η 1 +η 2 =1, ξ 1 ,ξ 2 represents the adjustable parameters of the design; E pu ,E qu Indicates the intermediate parameter; U k (t),U(t),U(t k -τ) represent the rudder angle / speed of the target unmanned ship when it is triggered at time k, the continuous output signal of the rudder angle / speed of the target unmanned ship, and the continuous output signal of the rudder angle / speed of the target unmanned ship at time t k -Rudder angle / speed when triggered at time τ;

[0190] S712: According to the state change amount E u Design a state memory type event trigger mechanism, whose expression is t k+1 = inf{t>t k ||E u (t)|≥ξ 3 |U(t)|+ξ 4}

[0191] Where: t k+1 ,tk represents the time parameter; ξ 3 ,ξ 4 represents the adjustable parameters of the design;

[0192] S713: Define the second intermediate parameter Ω i and

[0193] And according to the second intermediate parameter combined with the state memory type event trigger mechanism, the control input of the target unmanned ship is obtained, and its expression is:

[0194]

[0195] Where: 1 ,λ 2 represents the adjustable parameters of the design; Represents Ω i The estimated value of u represents the virtual control law of the longitudinal velocity of the target unmanned ship; i represents the intermediate parameter and

[0196] S72: Based on the dynamic model of the target unmanned ship and step S51, define the second intermediate parameter u e ,r e and u e =β u -u,r e =β r -r;

[0197] And according to the second intermediate parameter and the control input of the target unmanned ship, the third Lyapunov function V is constructed 3 , whose expression is

[0198]

[0199] Where: represents the design parameters of the third Lyapunov function; Represents the second intermediate parameter Ω u ,Ω r The estimation error of Represents χ u ,x r The estimation error of

[0200] Get the third Lyapunov function V 3 The derivative of Its expression is

[0201]

[0202] S73: In order to satisfy the derivative of the third Lyapunov function Stability, based on the reduced-order extended state observer with self-feedback compensation, the controller / adaptive law of the rudder angle / speed of the target unmanned ship is constructed, and its expression is

[0203]

[0204] Where: α n Indicates the speed of the target unmanned ship controller; α δ represents the rudder angle of the target unmanned ship controller; k u2 ,k r3 represents an adjustable design parameter; represents the estimated value of u,r by the reduced-order extended state observer compensated by self-feedback; Represents χ u ,x r An estimated value of represents the design parameters; Respectively represent Ω r ,Ω u Estimated value of Ω u (0),Ω y (0) respectively represent Ω r ,Ω u The initial value of represents the design parameter; y (0),x u (0) respectively represent χ y ,x u The initial value of express The first derivative of express The first derivative of

[0205] S8: According to the yaw angular velocity virtual control law / adaptive law, the rudder angle / rotation speed controller / adaptive law, the anti-deception attack control of the target unmanned ship is realized.

[0206] Compared with the prior art, the ship anti-cheating attack control method based on the self-feedback extended state observer disclosed in this embodiment has the following beneficial effects:

[0207] 1) The present invention aims at online compensation of deception attack signals. The proposed self-feedback extended state observer abandons the correction of the erroneous second-order position signal error, fundamentally avoiding the influence of deception attacks on error estimation, and greatly improving the observation capability of the extended state of the unmanned ship.

[0208] 2) The influence of the deception attack signal is decomposed into the coordinate system of the unmanned ship's appendage, and the yaw angular velocity virtual control law / adaptive law is obtained according to the constructed longitudinal velocity virtual control law and the deception attack compensation law. A set of adaptive parameters is designed for the rate of change of the target unmanned ship with respect to time, which can accurately compensate the control of the target unmanned ship under the deception attack.

[0209] 3) The L1 guidance principle of drones is improved and applied to the guidance of under-actuated unmanned ships, and the L2-DVS guidance strategy is constructed in combination with the DVS guidance principle. At the route generation level, it can avoid the problem of accumulation of calculation errors of the bow angle reference velocity signal over time, and produce a smoother reference path. At the level of guiding unmanned ships, when the distance between the real ship and the logical virtual ship is too large, the target unmanned ship can be guided by the virtual ship to approach the logical virtual ship in a smooth route.

[0210] 4) The state memory event trigger mechanism proposed in the present invention compensates for the problem that the traditional mixed threshold event trigger mechanism only considers the state of the previous moment and releases a small amount of control signal due to the reduced change rate before and after the peak by increasing the use of past control signal state values, and effectively reduces excessive wear of the actuator.

[0211] In order to verify the effectiveness of the method proposed in this embodiment, numerical simulation is performed through a simulation platform, and the results shown in the following figure are obtained. This embodiment mainly verifies the superiority of the algorithm of this embodiment from the following four aspects:

[0212] 1. Path tracking effect of unmanned ship under the influence of long-term large deception attack.

[0213] 2. The observation effect of the extended state observer proposed in this embodiment on the nonlinear part of the model and environmental disturbances in an attack environment.

[0214] 3. Whether the state memory event trigger mechanism proposed in this embodiment effectively reduces the wear of the actuator.

[0215] 4. Whether the adaptive compensation algorithm designed in this embodiment can effectively compensate for deception attacks.

[0216] 5. Whether the path generated by the L2-DVS guidance algorithm is smooth and easy to navigate.

[0217] In order to perform the path tracking task and determine the navigation path in this embodiment, the waypoints selected in this embodiment are as follows: 1 (150m,150m),W2(600m,200m),W 3 (900m,300m),W 4 (1350m,350m),W 5(1650m, 450m), the initial state of the unmanned ship, that is, its position in the geodetic coordinate system, initial heading angle and velocity components are set to: [x(0), y(0), ψ(0), u(0), v(0), r(0)] = [140m, 140m, -60°, 0m / s, 0m / s, 0m / s].

[0218] The following interference model is used to simulate random environmental interference;

[0219]

[0220] like Figures 4 to 5 As shown, in order to verify the effectiveness of the attack compensation effect of this embodiment, this embodiment applies a long-term continuous strong deception attack, and the attack model is shown in the following formula:

[0221] Ξ x =20+0.3cos(0.6t)

[0222] Ξ y =20+0.5cos(0.6t)

[0223] The attack time is 90s. It can be seen that the position of the ship has not deviated significantly, and the tracking effect is excellent. At the same time, it can be seen that the L2-DVS guidance strategy has a significant effect on the generation of a smooth path. Due to the good control effect, the DVS and LVS did not separate. Figure 8 It can be seen that the deception attack compensation method of this embodiment can respond quickly to the attack and has an excellent compensation effect.

[0224] like Figure 6 to Figure 7 The figure shows the observation effect of the reduced-order self-error feedback extended state observer of this embodiment. It can be seen that the observer can perform high-precision observation at the velocity layer, thereby ensuring accurate observation of the extended state.

[0225] like Fig. 9 As shown, this is an event trigger effect diagram, and the vertical axis represents the duration of a single trigger and the horizontal axis is the time axis. It can be seen that the release of the control command is effectively screened and the actuator wear is effectively alleviated.

[0226] The relevant documents involved in this embodiment are as follows:

[0227] [1]Zhang G, ShangX, LiuJ, et al. Improved iterative learningpath-followingcontrol for USVvia the potential-basedDVS guidance[J].OceanEngineering,2023,280

[0228] [2]Lekkas M A,Fossen I T.Integral LOS Path Following for Curved PathsBased on a Monotone CubicHermite SplineParametrization.[J].IEEETrans.Contr.Sys.Techn.,2014,22(6):2287-2301.

[0229] [3]Zhang G,Zhang C,Zhang X,et al.ESO-based path following control forunderactuated vehicles withthe safetypredictionobstacle avoidancemechanism[J].OceanEngineering,2019,188106259-106259.

[0230] [4]Ye H,Wu S,Liu W,et al.Adaptive neural synergetic heading controlfor USVs with unknown dynamicsanddisturbances[J].OceanEngineering,2024,300117438-.

[0231] [5]Zhang Z,Huang P,Gu H,et al.ESO-based and FTDO-based anti-swingcontrol for overhead craneswith externaldisturbance[J].MeasurementandControl,2025,58(1):50-59.

[0232] [6]Zhao Y,Yang X,Zhao Z.Pitch control for floating offshore windturbines via model-dependent extended state observer-based active disturbancerejection control[J].Ocean Engineering,2025,316119883-119883.

[0233] [7]Wang Z, Liu

[0234] [8]Mu

[0235] [9] Liang Wenxin, Song Shiwang, Zheng Yu, et al. Trajectory tracking control of UAV based on improved L1 guidance law under high sea conditions [J / OL]. Journal of Naval Aviation University, 1-13 [2025-01-03].

[0236]

[10] Xu B, Hu

[0237] 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 aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned 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 ship anti-cheating attack control method based on self-feedback extended state observer, characterized in that: The specific steps include: S1: Setting the guided virtual ship and dynamic virtual ship to construct the L2-DVS guidance strategy; S2: Obtain the kinematic model and nonlinear dynamic model of the target unmanned ship; S3: Construct an additional signal model for the deception attack of the unmanned ship, and obtain the ship position signal under the deception attack according to the kinematic model of the target unmanned ship; S4: Based on the kinematic model, according to the ship position signal and L2-DVS guidance strategy, the error model of the target unmanned ship in the attached coordinate system is obtained; According to the error model, the virtual control law of the longitudinal speed of the target unmanned ship and the deception attack compensation law of the yaw angle are constructed; S5: By introducing the dynamic surface control technology, the yaw rate virtual control law / adaptive law is obtained according to the longitudinal velocity virtual control law and the deception attack compensation law; S6: Based on the nonlinear dynamic model of the target unmanned ship, a reduced-order extended state observer with self-feedback compensation is constructed; S7: Construct a state memory type event trigger mechanism, and combine it with the reduced-order extended state observer with self-feedback compensation to obtain the controller / adaptive law of the rudder angle / speed of the target unmanned ship; S8: According to the yaw angular velocity virtual control law / adaptive law, the rudder angle / rotation speed controller / adaptive law, the anti-deception attack control of the target unmanned ship is realized.

2. According to claim 1, a ship anti-cheating attack control method based on self-feedback extended state observer is characterized in that: The L2-DVS guidance strategy constructed in S1 specifically sets a guidance virtual ship for obtaining a reference path of the target unmanned ship and a dynamic virtual ship for guiding the movement of the target unmanned ship; The expression of the guided virtual ship model is: The expression of the dynamic virtual ship model is: Where: x g ,y g ,ψ g They represent the longitudinal position, lateral position and heading angle of the guided virtual ship respectively; Respectively represent x g ,y g ,ψ g The first derivative of u g ,v g Respectively represent the longitudinal speed and lateral speed of the guided virtual ship; r d represents the yaw angular velocity of the guided virtual ship, that is, the reference yaw angular velocity of the target unmanned ship; x d ,y d They represent the longitudinal position and lateral position of the dynamic virtual ship, i.e., the position reference signal of the target unmanned ship; ψ rd ,u d ,v d They represent the yaw angle, longitudinal velocity and lateral velocity of the dynamic virtual ship respectively; and Represents x d ,y d The first derivative of d represents the yaw angle of the dynamic virtual ship, i.e., the yaw direction reference signal of the target unmanned ship; x, y represent the longitudinal position and lateral position of the target unmanned ship in the geodetic coordinate system respectively; Get the relative distance z between the target unmanned ship and the guided virtual ship e ; The relative distance z e The expression is Set the distance threshold L d , compare the relative distance z e With distance threshold L d size; If the relative distance z is confirmed e Less than the distance threshold L d That is z e <L d , then the position signal of the guided virtual ship is used as the position reference signal of the target unmanned ship, that is, x d =x g ,and d =and g If the relative distance z is confirmed e Greater than or equal to the distance threshold L d That is z e ≥L d , then according to the distance threshold L between the guided virtual ship and d , obtain the position reference signal of the target unmanned ship, that is x d =x g +L d cosψ d ,y d =y g +sinψ d 。 3. The ship anti-cheating attack control method based on self-feedback extended state observer according to claim 2 is characterized in that: The kinematic model and nonlinear dynamic model of the target unmanned ship obtained in S2; The expression of the kinematic model is The expression of the nonlinear dynamic model is Where: x, y, ψ represent the ship position and heading signal of the target unmanned ship respectively; u, v, r represent the longitudinal velocity, lateral velocity and yaw angular velocity of the target unmanned ship respectively; n, δ represent the control input provided by the host and the steering gear respectively; T u (·) and F r (·) represents the control gain function; d wu ,d wv ,d wr Respectively represent the interference items caused by wind, waves and currents in the ocean environment; m u ,m v ,m r ,d u1 ,d v1 ,d r1 ,d u2 ,d v2 ,d r2 ,d u3 ,d v3 ,d r3 represents the unknown parameters of the unmanned ship's inertia, hydrodynamic damping, and nonlinear damping terms; f u (υ),f v (υ),f r (υ) represents the nonlinear function of higher-order hydrodynamic effects.

4. The ship anti-cheating attack control method based on self-feedback extended state observer according to claim 3 is characterized in that: S3 specifically includes the following steps S31: Construct an additional signal model for the deception attack of unmanned ships, which is expressed as follows: Where: i represents the additional deception attack signal in the x,y direction of the target unmanned ship, and i = [x, y] T , a i ,b i , represents a non-zero constant of the design and a i =[a x ,a y ,a ψ ] T ,b i =[a x ,a y ,a ψ ] T , Indicates the use of trigonometric functions to simulate nonlinear functions with unknown parameters; S32: According to the kinematic model of the target unmanned ship combined with the deception attack signal of the deception attack additional signal model, the ship position signal under the deception attack is obtained, and its expression is:

5. The ship anti-cheating attack control method based on self-feedback extended state observer according to claim 4 is characterized in that: The S4 specifically includes the following steps S41: According to the ship position signal and the L2-DVS guidance strategy, the error model of the target unmanned ship in the attached coordinate system is obtained; The expression of the error model is: Where: x ,x y Denote the deception attack signal Ξ x With Ξ y The first derivative of x e ,y e ,ψ e Indicates the longitudinal position error, lateral position error and heading angle error of the target unmanned ship; S42: Construct the first Lyapunov function V1 according to the error model to obtain the derivative of the first Lyapunov function The first Lyapunov function V1 is The derivative of the first Lyapunov function is for S43: In order to satisfy the derivative of the first Lyapunov function Stability, construct the virtual control law of the longitudinal speed of the target unmanned ship and the deception attack compensation law of the yaw angle, and its expression is Where: Respectively represent χ x ,x y The estimated value of u represents the virtual control law of longitudinal velocity; K represents the deception attack compensation law of yaw angle; u ,K r1 Represents design parameters.

6. The ship anti-cheating attack control method based on self-feedback extended state observer according to claim 5 is characterized in that: S5 specifically includes the following steps S51: By introducing the dynamic surface control technology, the longitudinal speed virtual control law and the deception attack compensation law are filtered and the expression is: α i (0)＝β i (0) i=r,u,ψ e Where: β i represents a first-order filter; α i It is a general term for the virtual control law of longitudinal velocity, the deception attack compensation law and the virtual control law of yaw angular velocity; τ i represents the time constant of the first-order filter; α i (0), β i (0) indicates α i With β i The initial value of S52: Define the first intermediate parameter according to step S51 And y u =α u -β u , And construct the second Lyapunov function V2 based on the intermediate parameter, and its expression is Where: represents the adjustable positive design parameter of the second Lyapunov function; Represents χ x , χ y The estimation error of Get the derivative of the second Lyapunov function V2 Its expression is Where: τ u , represents the time constant of a first-order filter; and represents the design function with respect to the derivative of the second Lyapunov function V2; S53: In order to satisfy the derivative of the second Lyapunov function Stability, construct the virtual control law / adaptive law of the yaw angular velocity of the target unmanned ship, and its expression is Where: α r represents the virtual control law of yaw rate; K r represents the design parameters; express The first derivative of r It means that the virtual control law α of the yaw angular velocity is controlled by the dynamic surface control technology. r Time constant for first-order filtering.

7. A ship anti-cheating attack control method based on self-feedback extended state observer according to claim 6, characterized in that: The reduced-order extended state observer of the self-feedback compensation constructed in S6 is expressed as Where: G = [T u , F r T , i = [u, r] T , U = [|n|n, δ] T ;​ L1, L2 represent the adjustable parameters of the design, J i represents the extended state of the extended state observer and E i represents the velocity observation error and represents the expansion state observation error and Indicates J i An estimated value of represents the estimated value of i.

8. The ship anti-cheating attack control method based on self-feedback extended state observer according to claim 7 is characterized in that: The S7 specifically comprises the following steps: S71: The constructed state memory event trigger mechanism includes: S711: Define the state change E of the target unmanned ship's rudder angle / speed u , whose expression is E u =η1E pu +η2E qu E pu =U k (t)-U(t) E qu =U k (t)-U(t k -τ) Where: η1, η2 represent positive real numbers and η1+η2=1, ξ1, ξ2 represent the adjustable parameters of the design; E pu , E qu Indicates the intermediate parameter; U k (t), U(t), U(t k -τ) represent the rudder angle / speed of the target unmanned ship when it is triggered at time k, the continuous output signal of the rudder angle / speed of the target unmanned ship, and the continuous output signal of the rudder angle / speed of the target unmanned ship at time t k -Rudder angle / speed when triggered at time τ; S712: According to the state change amount E u Design a state memory event trigger mechanism, the expression is t k+1 =inf{t>t k ||E u (t)|≥ξ3|U(t)|+ξ4} Where: t k+1 , t k represents the time parameter; ξ3, ξ4 represent the adjustable parameters of the design; S713: Define the second intermediate parameter Ω i and i = [u, r] T ; And according to the second intermediate parameter combined with the state memory type event trigger mechanism, the control input of the target unmanned ship is obtained, and its expression is: Where: λ1, λ2 represent the adjustable parameters of the design; Represents Ω i The estimated value of u represents the virtual control law of the longitudinal velocity of the target unmanned ship; i represents the intermediate parameter and i = [u, r] T ; S72: Based on the dynamic model of the target unmanned ship and step S51, define the second intermediate parameter u e , r e Andu e =β u -u,r e =β r -r; According to the second intermediate parameter and the control input of the target unmanned ship, the third Lyapunov function V3 is constructed, and its expression is: Where: represents the design parameters of the third Lyapunov function; Represents the second intermediate parameter Ω u ,Ω r The estimation error of Represents χ u , χ r The estimation error of Get the derivative of the third Lyapunov function V3 Its expression is S73: In order to satisfy the derivative of the third Lyapunov function Stability, based on the reduced-order extended state observer with self-feedback compensation, the controller / adaptive law of the rudder angle / speed of the target unmanned ship is constructed, and its expression is Where: α n Indicates the speed of the target unmanned ship controller; α δ represents the rudder angle of the target unmanned ship controller; k u2 , k r3 represents an adjustable design parameter; represents the estimated values ​​of u, r by the reduced-order extended state observer compensated by self-feedback; Represents χ u , χ r An estimated value of represents the design parameters; Respectively represent Ω r ,Ω u Estimated value of Ω u (0),Ω y (0) respectively represent Ω r ,Ω u The initial value of represents the design parameter; y (0), χ u (0) respectively represent χ y , χ u The initial value of express The first derivative of express The first derivative of .

Images