Ship air-sea cooperative task guidance and gain adaptive control method oriented to surround alert

Abstract

The invention discloses a ship air-sea cooperative task guidance and gain adaptive control method for surround alert. The method comprises the following steps: constructing a mixed-order ship-unmanned aerial vehicle nonlinear system model; setting a waypoint path planning reference path, and generating a ship reference signal; acquiring the real-time position of the ship and the surrounding radius of the unmanned aerial vehicle, and generating an unmanned aerial vehicle reference signal by combining the virtual unmanned aerial vehicle; defining a ship-unmanned aerial vehicle kinematics error according to the reference signal, and designing a virtual control law; a dynamic surface technology is introduced, a dynamic surface signal of the virtual control law is obtained, and a dynamic error is defined; a radial basis function neural network and a minimum learning parameter technology are used to carry out approximation on a nonlinear term of a ship-unmanned aerial vehicle system, an MLP technology is introduced to simplify external interference, a preset performance control law and an adaptive law of an actuator gain are designed, and path tracking control under an unmanned aerial vehicle surrounding warning task is realized; the method can guarantee the control precision, reduces the design complexity of the controller, gives full play to the maneuverability of the unmanned aerial vehicle, and achieves the purpose that a ship and the unmanned aerial vehicle execute a surrounding warning task.

Description

Technical Field

[0001] The present invention relates to the technical field of ship and unmanned aerial vehicle motion control, and in particular to a ship-air-sea collaborative mission guidance and gain adaptive control method for surround alert. Background Art

[0002] The path tracking control system executed by the ship-UAV collaborative system consists of three subsystems: guidance, control, and navigation. The guidance system can automatically construct a reference signal based on the positional relationship between the current attitude of the ship-UAV system and the desired path; the control system can achieve effective convergence by stabilizing the error between the current attitude and the reference signal; the navigation system can transmit the position and attitude information of the controlled object to the guidance system and control system through sensors.

[0003] Guidance and control are two important subsystems in path tracking control. In the existing research results, the most common LOS guidance is simple in calculation, but it is greatly affected by the disturbance of the ocean environment and is not suitable for guidance with large errors. Vector field guidance requires more accurate data of ship or UAV model parameters and cannot be applied to different environments. Traditional virtual guidance (such as LVS guidance) mostly relies on time to divide the path into straight segments and turning segments for separate calculations. In the control system, there is no complete control theory for the heterogeneous cooperative control of ships and UAVs. Although the 3D mapping guidance algorithm constructs the guidance framework of the UAV-ship cooperative system, due to the limitation of the consistency of the ship and the ship, that is, the trajectory of the UAV and the ship is consistent, the algorithm is not suitable for the aircraft / ship circumvention warning task. In addition, most of the literature simply regards the actuator gain as a constant calculation, but in engineering practice, due to the change of the ship or UAV's own parameters and the different environments that affect the actuator gain, it is often difficult to obtain an accurate actuator gain.

[0004] Based on the above analysis, the traditional guidance and control method directly applied to the ship-UAV circumvention warning mission has the following two main defects:

[0005] 1) In the past, drones often assisted ships in navigation by synchronizing their positions with those of the ship. In this way, drones could also provide ships with information about the surrounding environment and warn of potential dangers. However, this collaborative approach could not effectively utilize the advantages of drones’ high maneuverability. In the surround alert, the drones would move around the ship at a uniform angular velocity while accompanying the ship. This would give full play to the advantages of the drone / ship collaborative system and expand the drone’s monitoring range. However, the use of surround alert tasks requires a specific guidance method, and the traditional consistent guidance method for the ship and the aircraft is obviously not applicable. In addition, traditional virtual guidance (such as LVS guidance) requires first calculating the time of the straight segment and the turning segment, and then analyzing the guidance reference signal. In navigation practice, this guidance strategy will produce cumulative errors in position and attitude at the switching point (straight segment-turning segment). For navigation situations with many waypoints (such as narrow waterways and island and reef waters), the accuracy of path tracking control will be affected.

[0006] 2) In engineering practice, the actuator gains of ships and drones are often difficult to measure and will change with use (e.g. due to wear or load) and the environment. Although it may be a valid assumption to treat the actuator gain as a known constant in some simplified analysis, in actual control system design, considering the change and dynamic characteristics of the actuator gain usually leads to better performance and system robustness. To solve this problem, designing adaptive gains is a method commonly used in ship control, but due to the strong coupling of drone models, this method cannot be directly applied to drone control. Summary of the invention

[0007] The present invention provides a ship-air-sea collaborative mission guidance and gain adaptive control method for surround warning, so as to overcome the technical problems that the existing ship-UAV collaborative system cannot be applied to UAVs for surround warning missions and cannot use adaptive gain instead of UAV actuator gain for control.

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

[0009] A method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert, comprising:

[0010] S1: Construct a mixed-order ship-UAV nonlinear system model as the controlled object of the control law designed in the subsequent steps;

[0011] S2: Setting a waypoint path, using a virtual ship to plan a reference path according to the waypoint path, and generating a ship reference signal; obtaining the real-time position of the ship and the circling radius of the drone, and generating a drone reference signal in combination with the virtual drone; the ship reference signal includes a ship position reference signal and a ship attitude reference signal;

[0012] S3: Based on the ship reference signal and the UAV reference signal, define the ship-UAV kinematic error, and design the corresponding virtual control law to eliminate the kinematic error of the ship-UAV nonlinear system model; the kinematic error includes the position error and attitude error of the ship-UAV;

[0013] S4: Introducing dynamic surface technology, obtaining the dynamic surface signal of the virtual control law, and defining a dynamic error according to the dynamic surface signal; the dynamic error includes a speed error and an angular velocity error of the ship-UAV;

[0014] S5: using radial basis function neural network and minimum learning parameter technology to simultaneously approximate the nonlinear terms of the ship-UAV nonlinear system model, and introducing MLP technology to simplify external interference;

[0015] S6: Using the approximated simplified ship-UAV nonlinear system model and the dynamic error, design a preset performance control law and an adaptive law of actuator gain;

[0016] S7: Path tracking control of the UAV in the surround warning mission is realized according to the preset performance control law and the adaptive law of the actuator gain.

[0017] Furthermore, a mixed-order ship-UAV nonlinear system model is constructed, as shown in formulas (1)-(5):

[0018]

[0019] In formula (1), H1=[x s ,y s ,ψ s ] T It represents the ship's forward movement, drift displacement and bow rolling angle, H2 = [x a ,y a ,z a ] T Indicates the forward, drift, and heave displacement of the drone, H3=[φ a ,θ a ,ψ a ] T Indicates the roll, pitch and yaw angles of the drone; I1 = [u s ,v s ,r s ] T Indicates the forward, drifting and pitching speed of the ship, I2=[u x ,u y ,u z ] T and I3=[p a ,q a ,r a ]T represents the speed and angular velocity of the drone along the ox, oy, oz axes in the coordinate system with the fuselage as the origin; F1 = [f u ,f v ,f r ] T , F2=[f x ,f y ,f z ] T and F3 = [f φ ,f θ ,f ψ ] T It represents the nonlinear term of the ship-UAV system, and its specific expansion is shown in formula (3); J1 in formula (2) is an intermediate variable;

[0020] M1=diag[m u ,m v ,m r ] represents the additional mass of the ship in the forward, drifting and bow rolling directions, M2 = diag[m a ,m a ,m a ] represents the mass of the drone, M3=diag[I xx ,I yy ,I zz ] represents the rotational inertia of the drone along the ox, oy, oz axis; C1=diag[C u ,0,C r ],C2=diag[k t ,k t ,k t ],C3=diag[k t ,k t ,k d ], are all unknown system actuator gain matrices; among them, C u ,C r is the ship actuator gain, k t ,k d is the UAV actuator gain; T1=[n|n|,0,δ] T represents the ship control input, n and δ represent the ship propeller speed and rudder angle respectively, T2 and T3 are the UAV control inputs, which are specifically expanded as shown in formulas (4) and (5), where Ω is the UAV propeller rotor speed;

[0021] D1=[d wu ,d wv ,d wr ] T , D2=[d wx ,d wy ,d wz ]T and D3=[d wφ ,d wθ ,d wψ ] T Respectively represent the external interference force / torque on the ship in the forward, drifting, and pitching directions and the UAV in the forward, drifting, heave, roll, pitch, and pitching directions; G = [0, 0, g] T represents the acceleration due to gravity;

[0022] In formula (3), R2 = diag[R x ,R y ,R z ], R3 = diag[l,l,1], both represent the fixed actuator gain matrix, where l is the distance from the center of the drone to each rotor; d ii1 ,d ii2 ,d ii3 ,ii=u,v,r represents the nonlinear damping term of the model; k dx ,k dy ,k dz It represents the rotational drag coefficient of the drone along the ox, oy, oz axis;

[0023] f u , f v , f r They represent the nonlinear terms of the ship in the forward, drift and bow rolling directions, respectively, and f x , f y , f z They represent the nonlinear terms of the UAV in the forward, drift, and heave directions, respectively, and f θ ,f θ ,f ψ They represent the nonlinear terms of the UAV in the roll, pitch and yaw directions, R x ,R y ,R z Respectively represent the gain coefficients of the drone on ox, oy, and oz.

[0024] Further, a waypoint path is set, a reference path is planned according to the waypoint path using a virtual ship, and a ship reference signal is generated; the real-time position of the ship and the circling radius of the drone are obtained, and a drone reference signal is generated in combination with the virtual drone, including:

[0025] S21, set the waypoint path and the ship turning radius, and select the target point P according to the set turning radius before and after each waypoint except the starting point and the end point inWk and P outWk ;

[0026] S22, using a virtual ship to plan a reference path according to the waypoint path, as shown in formula (6),

[0027]

[0028] In the formula, x sd ,y sd ,ψ sd Indicates the forward, drifting distance and heading angle of the virtual ship L1, that is, the ship position reference signal, u sd ,r sd represents the forward speed and yaw angular velocity of L1 ship, u sd According to the actual situation, set sd The calculation formula is shown in formula (7).

[0029]

[0030] Where, L 1s is the distance from L1 ship to the target point, η s =ψ sd -ψ L1s , ψ L1s is the azimuth from L1 ship to the target point, η s Indicates the deviation angle between L1 ship and target point; a ls represents the lateral acceleration of the L1 ship;

[0031] The azimuth from the actual ship to the L1 ship is shown in formula (8):

[0032]

[0033] In the formula, ψ r Indicates the azimuth from the actual ship to the L1 ship, i.e., the ship attitude reference signal, x se =x sd -x s ,y se =y sd -y s , indicating the forward and drift distance error of the ship;

[0034] S23, obtaining the real-time position of the ship and the circling radius of the drone;

[0035] S24, using the virtual drone to obtain the drone's reference signal, as shown in formula (9),

[0036]

[0037] In the formula, x ad ,y ad ,ψ ad represents the forward and drift distances and heading angles of the virtual UAV, i.e., the reference signal of the UAV, T a is the set UAV orbit period, LR is the set UAV orbit radius, η a is the azimuth from the ship to the virtual drone, pp(0) its initial value, x ae =x ad -x a ,y ae =y ad -y a , is the UAV's forward and lateral distance error, z ad is the desired rise and fall altitude of the drone.

[0038] Furthermore, according to the ship reference signal and the UAV reference signal, the ship-UAV kinematic error is defined, and the corresponding virtual control law is designed, including:

[0039] S31. Define the ship-UAV position error based on the ship reference signal and the UAV reference signal, as shown in formula (10):

[0040]

[0041] In the formula, z se represents the ship position error, z ae Indicates the heave distance error of the drone;

[0042] S32, deriving the position error of the ship-UAV, as shown in formula (11),

[0043]

[0044] In the formula, ψ se =ψ r -ψ s is the ship’s heading angle error;

[0045] S33, defining a preset performance function of the ship-UAV position error, as shown in formula (12),

[0046]

[0047] In the formula, ρ i ,i=u,x,y,z, is the preset performance function of position error, ρ i0 and ρ i∞ is i The initial value and final convergence value of k αi is a constant;

[0048] Define the conversion error of the ship-UAV position error as shown in formula (13):

[0049]

[0050] In the formula, σ i ,i=u,x,y,z represents the conversion error of position error;

[0051] S34. Design the virtual control law of the position error of the ship-UAV, as shown in formula (14):

[0052]

[0053] In the formula, k αu ,k αx ,k αy ,k αz is the virtual control law α u ,α x ,α y ,α z design parameters.

[0054] Furthermore, the kinematic error of the ship-UAV is defined and the corresponding virtual control law is designed, which also includes:

[0055] S35. Define the ship-UAV attitude error according to the ship reference signal and the UAV reference signal, as shown in formula (15):

[0056]

[0057] In the formula, ψ se Indicates the ship's heading angle error, φ ae represents the roll angle error of the UAV, θ ae represents the pitch angle error of the UAV, ψ ae Represents the UAV heading angle error; φ ad represents the reference roll angle of the UAV, θ ad Indicates the reference roll and pitch angles of the drone;

[0058] S36, deriving the attitude error of the ship-UAV, as shown in formula (16),

[0059]

[0060] S37, defining a preset performance function of the ship-UAV attitude error, as shown in formula (17),

[0061] ρ j (t)=(ρ j0 -ρ j∞ ) -kjt +ρ j∞ (17)

[0062] In the formula, ρ j ,j=r,φ,θ,ψ, is the preset performance function of attitude error, ρ j0 and ρj∞ is j The initial value and final convergence value of k αj is a constant;

[0063] Define the conversion error of the ship-UAV attitude error as shown in formula (18):

[0064]

[0065] In the formula, σ j ,j=r,φ,θ,ψ, represents the attitude error conversion error;

[0066] S38. Design the virtual control law of the ship-UAV attitude error, as shown in formula (19):

[0067]

[0068] In the formula, k αr ,k αφ ,k αθ ,k αψ is the virtual control law of attitude error α r ,α φ ,α θ ,α ψ control parameters.

[0069] Furthermore, the dynamic surface technology is introduced to obtain the dynamic surface signal of the virtual control law, and the dynamic error is defined according to the dynamic surface signal, including:

[0070] S41. Introduce dynamic surface technology, reduce the derivative of the virtual control law of the position error and the derivative of the virtual control law of the attitude error, and obtain the dynamic surface error of the position error and the dynamic surface error of the attitude error, as shown in formulas (20) and (21).

[0071]

[0072] In formula (20), β i is α i The dynamic surface signal, ∈ i is the time constant, and the dynamic surface error of the position error is y i =α i -β i , α i A virtual control law representing the position error;

[0073] In formula (21), β j is α j The dynamic surface signal, ∈ j is the time constant, and the dynamic surface error of the attitude error is y j =α j-β j , α j A virtual control law representing the attitude error;

[0074] S42. Define the ship-UAV velocity error and take its derivative, as shown in formulas (22) and (23):

[0075]

[0076] In the formula, u se ,u xe ,u ye ,u ze represents the ship's forward movement, the UAV's forward movement, drift and heave speed errors, B2 = [β x ,β y ,β z ] T Dynamic surface signal matrix representing position error;

[0077] Define the angular velocity error between the ship and the UAV and take its derivative, as shown in equations (24) and (25):

[0078]

[0079] In the formula, r se ,p ae ,q ae ,r ae Indicates the angular velocity error of the ship's bow roll, the UAV's roll, pitch and bow roll; B3 = [β φ ,β θ ,β ψ ] T , represents the dynamic surface signal matrix of the attitude error.

[0080] Furthermore, the radial basis function neural network and the minimum learning parameter technology are used to simultaneously approximate the nonlinear terms of the ship-UAV nonlinear system model, and the MLP technology is introduced to simplify the external interference, as shown in formulas (26) and (27),

[0081] f i -d wi ≤Θ i1 Θ i2 -c i S i W i (26)

[0082] f j -d wj ≤Θ j1 Θ j2 -c j S j W j(27)

[0083] In the formula, f i and f j represents the nonlinear term of the ship-UAV nonlinear system model, Θ i2 =1+‖S i ‖+1,Θ j2 =1+‖S j ‖+1,Θ i1 ,Θ j1 ,Θ i2 ,Θ j2 It is the intermediate variable in the minimum learning parameter technique;

[0084] A i represents f i The neural network weight update law, A j represents f j The neural network weight update law, c i ,c j A i ,A j The norm of S i and S j represents the Gaussian function, ε i and ε j represents the approximation error, is the approximation error ε i The maximum value of is the approximation error ε j The maximum value, d mi It is the external environment interference wi The maximum value, d mj It is the external environment interference wj The maximum value of W i ,W j represents the intermediate variable, as shown in formula (28),

[0085]

[0086] Furthermore, the ship-UAV nonlinear system model after approximation and simplification and the dynamic error are used to design a preset performance control law and an adaptive law of actuator gain, including:

[0087] S61, obtaining the actual control input, and converting the actual control input into a preset performance control law and an adaptive law of actuator gain, as shown in formula (29),

[0088]

[0089] In the formula, τ l,l=u,r,x,y,z,φ,λ,ψ are the preset performance control laws of the ship's forward direction, bow angle, UAV's forward direction, drift, heave direction, roll, pitch and bow angle; n represents the propeller speed of the ship; δ represents the rudder angle of the ship;

[0090] λ2=[λ kt ,λ kt ,λ kt ] T ,λ3=[λ kt ,λ kt ,λ kd ] T , They are λ cu ,λ cr ,λ kt ,λ kd The estimated value of cu ,λ cr ,λ kt ,λ kd They represent the adaptive parameters of the actuator gain, i.e., the adaptive law, and represents the matrix of actuator gains, τ2 and τ3 represent the preset performance control law matrices;

[0091] S62. Using the simplified ship-UAV nonlinear system model, formulas (23) and (29), the MLP technology, coupling gain adaptation technology and backstepping method are introduced to design the preset performance control law of the speed error and the adaptive law of the actuator gain, as shown in formulas (30)-(32).

[0092]

[0093]

[0094] In the formula, k i ,Γ Cu1 ,Γ Cu2 , is the design parameter, ω i represents the adaptive parameter introduced by the minimum learning parameter, Yes i An estimated value of Ψ i Represents ω i The robust damping term of i1 ,ρ i2 is a constant; τ u ,τ x ,τ y ,τz represents the preset performance control law for the speed error, represents the derivative of the adaptive parameter;

[0095] S63. Using the simplified ship-UAV nonlinear system model, formulas (25) and (29), the MLP technology, coupling gain adaptation technology and backstepping method are introduced to design the preset performance control law of the angular velocity error and the adaptive law of the actuator gain, as shown in formulas (33)-(35).

[0096]

[0097] In the formula, k j ,Γ Cr1 ,Γ Cr2 ,Γ kt1 ,Γ kt2 ,Γ kd1 ,Γ kd2 , is the design parameter, Yes j The estimated value of ω j The robust damping term of j1 ,ρ j2 is a constant; τ j Denotes the preset performance control law for the angular velocity error, ω j Adaptive parameters introduced for the minimum learning parameter technique, The derivative of the adaptive parameter representing the actuator gain.

[0098] Furthermore, the nonlinear decoupling technology is used to solve formula (30) to obtain the reference roll angle and pitch angle of the UAV, as shown in formula (36):

[0099]

[0100] In the formula, φ ad represents the reference roll angle of the UAV, θ ad Indicates the reference pitch angle of the drone.

[0101] Beneficial effects: The present invention provides a ship-air-sea collaborative task guidance and gain adaptive control method for surround alert, which obtains the reference signal of the ship and the reference signal of the UAV for surround alert. Compared with the traditional UAV fixed above or in front of the ship, the surround alert can give full play to the maneuverability of the UAV, and has the advantages of not being easily interfered, having a large alert range, and having no blind spots in vision.

[0102] Considering the maneuverability of the ship-UAV heterogeneous system, a virtual control law of the ship-UAV kinematic error is designed. The radial basis function neural network and the minimum learning parameter technology are combined to simultaneously perform online approximation of the model structure uncertainties of the ship-UAV collaborative system and the external environmental disturbances. In addition, considering the problem of actuator gain uncertainty in actual navigation, an adaptive law of actuator gain and a preset performance control law are designed. While ensuring the control accuracy, the complexity of the controller design is reduced, thereby achieving the goal of the ship and UAV performing the surround surveillance mission. BRIEF DESCRIPTION OF THE DRAWINGS

[0103] 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.

[0104] Figure 1 A method flow chart of the method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert provided by the present invention;

[0105] Figure 2 A framework diagram for generating guidance reference signals using L1 ships;

[0106] Figure 3 To generate a guidance reference signal framework diagram using a virtual UAV;

[0107] Figure 4 It is a tracking trajectory diagram of the ship-UAV surrounding warning path of the present invention;

[0108] Figure 5 It is a schematic diagram of the propeller speed and rudder angle of the ship of the present invention;

[0109] Figure 6 Schematic diagram of the rotor speed of the UAV of the present invention;

[0110] Figure 7 Schematic diagram of the ship position error and attitude error of the present invention;

[0111] Figure 8 is a schematic diagram of the position error of the UAV of the present invention;

[0112] Fig. 9 Schematic diagram of the attitude error of the UAV of the present invention. DETAILED DESCRIPTION

[0113] 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.

[0114] This embodiment provides a method for ship-air-sea collaborative mission guidance and gain adaptive control for surround warning, such as Figure 1 As shown, including:

[0115] A method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert, comprising:

[0116] S1: Construct a mixed-order ship-UAV nonlinear system model as the controlled object of the control law designed in the subsequent steps;

[0117] S2: Setting a waypoint path, using a virtual ship to plan a reference path according to the waypoint path, and generating a ship reference signal; obtaining the real-time position of the ship and the circling radius of the drone, and generating a drone reference signal in combination with the virtual drone; the ship reference signal includes a ship position reference signal and a ship attitude reference signal;

[0118] S3: Based on the ship reference signal and the UAV reference signal, define the ship-UAV kinematic error, and design the corresponding virtual control law to eliminate the kinematic error of the ship-UAV nonlinear system model; the kinematic error includes the position error and attitude error of the ship-UAV;

[0119] S4: Introducing dynamic surface technology, obtaining the dynamic surface signal of the virtual control law, and defining a dynamic error according to the dynamic surface signal; the dynamic error includes a speed error and an angular velocity error of the ship-UAV;

[0120] S5: using radial basis function neural network and minimum learning parameter technology to simultaneously approximate the nonlinear terms of the ship-UAV nonlinear system model, and introducing MLP technology to simplify external interference;

[0121] S6: Using the approximated simplified ship-UAV nonlinear system model and the dynamic error, design a preset performance control law and an adaptive law of actuator gain;

[0122] S7: Path tracking control of the UAV in the surround warning mission is realized according to the preset performance control law and the adaptive law of the actuator gain.

[0123] Specifically, firstly, a mixed-order ship-UAV nonlinear system model is constructed as the controlled object of the control law designed in the subsequent steps, so as to prepare for the approximation of nonlinear terms in the subsequent steps and then design of preset performance control laws and adaptive laws; secondly, a waypoint path is set, and a virtual ship is used to plan a reference path according to the waypoint path, and a ship reference signal is generated; the real-time position of the ship and the UAV's orbiting radius are obtained, and a UAV reference signal is generated in combination with the virtual UAV to realize the UAV's orbiting warning mission, meet the requirements of navigation practice, and have higher guidance accuracy; based on the ship reference signal and the UAV reference signal, the kinematic error of the ship-UAV is defined, and the corresponding virtual control law is designed to eliminate the kinematic error of the ship-UAV nonlinear system model, so as to realize the precise tracking of the ship-UAV and return the path to the reference path. ; Secondly, the dynamic surface technology is introduced to obtain the dynamic surface signal of the virtual control law, and the dynamic error is defined according to the dynamic surface signal, which can reduce the load in the calculation process and improve the processing efficiency; the radial basis function neural network and the minimum learning parameter technology are used to simultaneously approximate the nonlinear terms of the ship-UAV nonlinear system model, and the MLP technology is introduced to simplify the external interference, which can reduce the adaptive parameters in the controller, which is beneficial to the maneuverability of the heterogeneous system and improve the efficiency of the UAV's coordinated surveillance of ships; finally, the ship-UAV nonlinear system model after approximation and simplification and the dynamic error are used to design the preset performance control law and the adaptive law of the actuator gain, so as to achieve the control effect of controlling the error within the budget range when the actuator gain is uncertain, and realize the path tracking control of the UAV under the surrounding surveillance task.

[0124] In a specific embodiment, a mixed-order ship-UAV nonlinear system model is constructed as the controlled object of the control law designed in the subsequent steps. The model is shown in formulas (37)-(41):

[0125]

[0126]

[0127] In formula (37), H1=[x s ,y s ,ψ s ] T It represents the ship's forward movement, drift displacement and bow rolling angle, H2 = [x a ,y a ,z a ] T Indicates the forward, drift, and heave displacement of the drone, H3=[φ a ,θ a ,ψ a ] TIndicates the roll, pitch and yaw angles of the drone; I1 = [u s ,v s ,r s ] T Indicates the forward, drifting and pitching speed of the ship, I2=[u x ,u y ,u z ] T and I3=[p a ,q a ,r a ] T represents the speed and angular velocity of the drone along the ox, oy, oz axes in the coordinate system with the fuselage as the origin; F1 = [f u ,f v ,f r ] T , F2=[f x ,f y ,f z ] T and F3 = [f φ ,f θ ,f ψ ] T represents the nonlinear term of the ship-UAV system, and the specific expansion is shown in formula (39); formula (38) J1 is an intermediate variable;

[0128] M1=diag[m u ,m v ,m r ] represents the additional mass of the ship in the forward, drifting and bow rolling directions, M2 = diag[m a ,m a ,m a ] represents the mass of the drone, M3=diag[I xx ,I yy ,I zz ] represents the rotational inertia of the drone along the ox, oy, oz axis; C1=diag[C u ,0,C r ],C2=diag[k t ,k t ,k t ],C3=diag[k t ,k t ,k d ], are all unknown system actuator gain matrices; among them, C u ,C r is the ship actuator gain, k t ,k d for the UAV actuator gains; in engineering practice, they are often uncertain;

[0129] T1=[n|n|,0,δ] T represents the ship control input, n and δ represent the ship propeller speed and rudder angle respectively, T2 and T3 are the UAV control inputs, and the specific expansion is shown in formulas (40) and (41), where Ω is the UAV propeller rotor speed;

[0130] D1=[d wu ,d wv ,d wr ] T , D2=[d wx ,d wy ,d wz ] T and D3=[d wφ ,d wθ ,d wψ ] T Respectively represent the external interference force / torque on the ship in the forward, drifting, and bowing directions and the UAV in the forward, drifting, heave, roll, pitch, and bowing directions. Here, it is assumed that the interference has a maximum value and supremum; G = [0,0,g] T represents the acceleration due to gravity;

[0131] In formula (39), R2 = diag[R x ,R y ,R z ], R3 = diag[l,l,1], both represent the fixed actuator gain matrix, where l is the distance from the center of the drone to each rotor; d ii1 ,d ii2 ,d ii3 ,ii=u,v,r represents the nonlinear damping term of the model; k dx ,k dy ,k dz It represents the rotational drag coefficient of the drone along the ox, oy, oz axis;

[0132] f u , f v , f r They represent the nonlinear terms of the ship in the forward, drift and bow rolling directions, respectively, and f x , f y , f z They represent the nonlinear terms of the UAV in the forward, drift, and heave directions, respectively, and f φ ,f θ ,f ψ They represent the nonlinear terms of the UAV in the roll, pitch and yaw directions, R x ,R y ,R z Respectively represent the gain coefficients of the drone on ox, oy, and oz.

[0133] In this scheme, a mixed-order ship-UAV nonlinear system model is constructed as the controlled object of the control law designed in the subsequent steps. At the same time, the nonlinear terms are obtained to approximate the nonlinear terms in the subsequent steps, and then prepare for the design of preset performance control laws and adaptive laws.

[0134] In a specific embodiment, a waypoint path is set, a reference path is planned according to the waypoint path using a virtual ship, and a ship reference signal is generated; the real-time position of the ship and the circling radius of the drone are obtained, and a solution for generating a drone reference signal in combination with the virtual drone is:

[0135] S21, such as Figure 2 As shown in the figure, the virtual guidance generates a ship reference signal which needs to be processed by the virtual ship L1 and generate a smooth reference path. First, a waypoint path consisting of several waypoints and a ship turning radius are set. The target point R is selected according to the set turning radius before and after each waypoint except the starting point and the end point. inWk and P outWk , when L1 ship reaches a target point, it switches to the next target point tracking; among them, the starting point does not select the target point, and the end point is the target point selected by the previous waypoint; in the figure, W k-1 ,W k ,W k+1 are three consecutive waypoints, waypoint W k Select the target point P according to the set turning radius inWk and P outWk , L1 ship is now at P inWk is the target point, and after tracking it, it switches to the next target point. Wk and Wk It is W k The coordinates of x target and target is the target point P at this time inWk The coordinates of L 1s is the distance from L1 ship to the target point, η s =ψ sd -ψ L1s , ψ L1s is the azimuth from L1 ship to the target point, η s represents the deviation angle between L1 ship and target point, ψ sd Indicates the heading angle of L1 ship.

[0136] S22, using a virtual ship to plan a reference path according to the waypoint path, as shown in formula (42),

[0137]

[0138] In the formula, x sd ,ysd ,ψ sd Indicates the forward, drifting distance and heading angle of the virtual ship L1, that is, the ship position reference signal, u sd ,r sd represents the forward speed and yaw angular velocity of L1 ship, u sd According to the actual situation, set sd The calculation formula is shown in formula (44),

[0139]

[0140] Where, L 1s is the distance from L1 ship to the target point, η s =ψ sd -ψ L1s , ψ L1s is the azimuth from L1 ship to the target point, η s Indicates the deviation angle between L1 ship and target point; a ls represents the lateral acceleration of the L1 ship;

[0141] The azimuth from the actual ship to the L1 ship is shown in formula (45):

[0142]

[0143] In the formula, ψ r Indicates the azimuth from the actual ship to the L1 ship, i.e., the ship attitude reference signal, x se =x sd -x s ,y se =y sd -y s , indicating the forward and drift distance error of the ship;

[0144] S23, such as Figure 3 As shown in the figure, the real-time position of the ship and the orbiting radius of the drone are obtained. The three-dimensional diagram on the left side of the figure is the motion trajectory of the ship-drone system within one orbiting cycle of the drone. The upper right corner is a diagram of the drone state variables, and the lower right corner is a top-down plan view of the trajectory diagram of the ship-drone system within one orbiting cycle of the drone. R is the set UAV orbit radius, η a is the azimuth from the ship to the virtual drone;

[0145] S24. Use the virtual drone to obtain the reference signal of the drone, as shown in formula (46),

[0146]

[0147] In the formula, x ad ,y ad ,ψad represents the forward and drift distances and heading angles of the virtual UAV, i.e., the reference signal of the UAV, T a is the drone orbit period set by the user, L R is the orbit radius of the drone set by the user, η a is the azimuth from the ship to the virtual drone, pp(0) its initial value, x ae =x ad -x a ,y ae =y ad -y a , is the UAV's forward and lateral distance error, z ad It is the expected rise and fall height of the drone, and the specific value is set by the user.

[0148] In this solution, a virtual L1 ship can be relied on to plan a reference path and generate a guidance signal according to the set waypoint path. The reference signal generated by the L1 ship is based on the distance between the ship and the target point rather than the time, which better meets the requirements of navigation practice and makes the guidance accuracy higher. In order to ensure that the drone can always maintain a suitable distance from the ship during navigation (meeting the warning range and not exceeding the communication distance), the reference signal of the drone is calculated by the real-time position of the ship and the set circling radius, so as to realize the drone's uniform angular speed periodic circling around the ship.

[0149] In a specific embodiment, according to the ship reference signal and the UAV reference signal, the kinematic error of the ship-UAV is defined, and the corresponding virtual control law is designed to eliminate the kinematic error of the ship-UAV nonlinear system model. The scheme is:

[0150] S31. Define the ship-UAV position error based on the ship reference signal and the UAV reference signal, as shown in formula (47):

[0151]

[0152] In the formula, z se represents the ship position error, z ae Indicates the heave distance error of the drone;

[0153] S32, deriving the position error of the ship-UAV, as shown in formula (48),

[0154]

[0155] In the formula, ψ se =ψ r -ψ s is the ship’s heading angle error;

[0156] S33, defining a preset performance function of the ship-UAV position error, as shown in formula (49),

[0157]

[0158] In the formula, ρ i ,i=u,x,y,z, is the preset performance function of position error, ρ i0 and ρ i∞ is i The initial value and final convergence value of k αi is a positive constant;

[0159] Define the conversion error of the ship-UAV position error as shown in formula (50):

[0160]

[0161] In the formula, σ i ,i=u,x,y,x represents the conversion error of position error;

[0162] S34, in order to stabilize the position error, a virtual control law of the position error of the ship-UAV is designed, as shown in formula (51):

[0163]

[0164] In the formula, k αu ,k αx ,k αy ,k αz is the virtual control law α u ,α x ,α y ,α z Positive design parameters;

[0165] S35. In order to control the current attitude of the ship-UAV to converge to the reference attitude, the attitude error of the ship-UAV is defined according to the ship reference signal and the UAV reference signal, as shown in formula (52):

[0166]

[0167] In the formula, ψ se Indicates the ship's heading angle error, φ ae represents the roll angle error of the UAV, θ ae represents the pitch angle error of the UAV, ψ ae Represents the UAV heading angle error; φ ad represents the reference roll angle of the UAV, θ ad Indicates the reference roll and pitch angles of the drone;

[0168] S36, deriving the attitude error of the ship-UAV, as shown in formula (53),

[0169]

[0170] S37, defining a preset performance function of the ship-UAV attitude error, as shown in formula (54),

[0171] ρ j (t)=(ρ j0 -ρ j∞ ) -kjt +ρ j∞ (54)

[0172] In the formula, ρ j ,j=r,φ,θ,ψ, is the preset performance function of attitude error, ρ j0 and ρ j∞ is j The initial value and final convergence value of k αj is a positive constant;

[0173] Define the conversion error of the ship-UAV attitude error as shown in formula (55):

[0174]

[0175] In the formula, σ j ,j=r,φ,θ,ψ, represents the attitude error conversion error;

[0176] S38. In order to stabilize the attitude error, a virtual control law for the attitude error of the ship-UAV is designed, as shown in formula (56):

[0177]

[0178] In the formula, k αr ,k αφ ,k αθ ,k αψ is the virtual control law α for the attitude error greater than zero r ,α φ ,α θ ,α ψ control parameters.

[0179] In this embodiment, the position error and attitude error of the ship are obtained, and a virtual control law is designed to eliminate the error, so that accurate tracking of the ship-UAV can be achieved and the path can be returned to the reference path.

[0180] In a specific embodiment, the dynamic surface technology is introduced to obtain the dynamic surface signal of the virtual control law, and the scheme for defining the dynamic error according to the dynamic surface signal is:

[0181] S41, the virtual controller will cause a large computational load problem in the subsequent derivation, so the dynamic surface technology is introduced to reduce the order of the derivative of the virtual control law of the position error and the derivative of the virtual control law of the attitude error to obtain the dynamic surface error of the position error and the dynamic surface error of the attitude error, as shown in formulas (57) and (58),

[0182]

[0183] In formula (57), β i is α i The dynamic surface signal, ∈ i is a time constant greater than zero, and the dynamic surface error of the position error is y i =α i -β i , α i A virtual control law representing the position error;

[0184] In formula (58), β j is α j The dynamic surface signal, ∈ j is a time constant greater than zero, and the dynamic surface error of the attitude error is y j =α j -β j , α j A virtual control law representing the attitude error;

[0185] S42. Define the ship-UAV velocity error and perform derivative, as shown in formulas (59) and (60),

[0186]

[0187] In the formula, u se ,u xe ,u ye ,u ze represents the ship's forward movement, the UAV's forward movement, drift and heave speed errors, B2 = [β x ,β y ,β z ] T Dynamic surface signal matrix representing position error;

[0188] Define the angular velocity error between the ship and the UAV and take its derivative, as shown in equations (61) and (62):

[0189]

[0190]

[0191] In the formula, r se ,pae ,q ae ,r ae Indicates the angular velocity error of the ship's bow roll, the UAV's roll, pitch and bow roll; B3 = [β φ ,β θ ,β ψ ] T , represents the dynamic surface signal matrix of the attitude error.

[0192] In this solution, dynamic surface technology is introduced to reduce the order of virtual control laws, which can reduce the load in the calculation process and improve processing efficiency.

[0193] In a specific embodiment, the nonlinear terms of the ship-UAV nonlinear system model are approximated by using radial basis function neural network and minimum learning parameter technology, and the MLP technology is introduced to simplify the external interference, as shown in formulas (63) and (64),

[0194] f i -d wi ≤Θ i1 Θ i2 -c i S i W i (63)

[0195] f j -d wj ≤Θ j1 Θ j2 -c j S j W j (64)

[0196] In the formula, f i and f j represents the nonlinear term of the ship-UAV nonlinear system model, Θ i2 =1+‖S i ‖+1,Θ j2 =1+‖S j ‖+1,Θ i1 ,Θ j1 ,Θ i2 ,Θ j2 It is the intermediate variable in the minimum learning parameter technique;

[0197] A i represents f i The neural network weight update law, A j represents f j The neural network weight update law, c i ,c j A i ,Aj The norm of S i and S j represents the Gaussian function, ε i and ε j represents the approximation error, is the approximation error ε i The maximum value of is the approximation error ε j The maximum value, d mi It is the external environment interference wi The maximum value, d mj It is the external environment interference wj The maximum value of W i ,W j represents the intermediate variable, as shown in formula (65),

[0198]

[0199] In this scheme, a neural network system and a minimum learning parameter strategy are combined to deal with the structural uncertainties and external environmental interference in the ship-UAV collaborative system. At the same time, a controller is designed for the ship-UAV system to reduce the adaptive parameters in the controller, which is beneficial to the maneuverability of the heterogeneous system and improves the efficiency of UAV collaborative surveillance of ships.

[0200] In a specific embodiment, the scheme for designing the preset performance control law and the adaptive law of the actuator gain using the approximated simplified ship-UAV nonlinear system model and the dynamic error is:

[0201] S61, obtaining the actual control input, and converting the actual control input into a preset performance control law and an adaptive law of actuator gain, as shown in formula (66),

[0202]

[0203] In the formula, τ l ,l=u,r,x,y,z,φ,θ,ψ are the preset performance control laws of the ship's forward direction, bow angle, UAV's forward direction, drift, heave direction, roll, pitch and bow angle; n represents the propeller speed of the ship; δ represents the rudder angle of the ship;

[0204] λ2=[λ kt ,λ kt ,λ kt ] T ,λ3=[λ kt ,λ kt ,λ kd ] T , They are λ cu ,λ cr ,λ kt ,λ kd The estimated value of cu ,λ cr ,λ kt ,λ kd They represent the adaptive parameters of the actuator gain, i.e., the adaptive law, and represents the matrix of actuator gains, τ2 and τ3 represent the preset performance control law matrices;

[0205] S62. Using the simplified ship-UAV nonlinear system model, formulas (60) and (66), the MLP technology, coupling gain adaptation technology and backstepping method are introduced to design the preset performance control law of the speed error and the adaptive law of the actuator gain, as shown in formulas (67)-(69).

[0206]

[0207] In the formula, k i ,Γ Cu1 ,Γ Cu2 , is a design parameter greater than zero, ω i represents the adaptive parameter introduced by the minimum learning parameter, Yes i An estimated value of Ψ i Represents ω i The robust damping term of i1 ,ρ i2 is a positive constant; τ u ,τ x ,τ y ,τ z represents the preset performance control law for the speed error, represents the derivative of the adaptive parameter;

[0208] The nonlinear decoupling technology is used to solve formula (67) to obtain the reference roll angle and pitch angle of the UAV, as shown in formula (70):

[0209]

[0210] In the formula, φ ad represents the reference roll angle of the UAV, θ ad Indicates the reference pitch angle of the drone;

[0211] S63. Using the approximated simplified ship-UAV nonlinear system model, formulas (62) and (66), the MLP technology, coupling gain adaptation technology and backstepping method are introduced to design the preset performance control law of the angular velocity error and the adaptive law of the actuator gain, as shown in formulas (71)-(73).

[0212]

[0213] In the formula, k j ,Γ Cr1 ,Γ Cr2 ,Γ kt1 ,Γ kt2 ,Γ kd1 ,Γ kd2 , is a design parameter greater than zero, Yes j The estimated value of ω j The robust damping term of j1 ,ρ j2 is a constant; τ j Denotes the preset performance control law for the angular velocity error, ω j Adaptive parameters introduced for the minimum learning parameter technique, The derivative of the adaptive parameter representing the actuator gain.

[0214] In this scheme, taking into account the problem of actuator gain uncertainty in actual navigation, a coupled gain adaptation is designed for all actuator gains of the ship-UAV system. Combined with the preset performance control, a preset performance robust adaptive control law and an adaptive law are designed to achieve the control effect of controlling the error within the budget range when the actuator gain is uncertain.

[0215] In order to perform the collaborative task of the drone surrounding the ship, this embodiment carried out a simulation experiment on the MATLAB platform, and selected a waypoint path consisting of 6 ship waypoints, where the drone's surrounding radius was 200m, the period was 50s, and the initial state of the controlled object was

[0216] [x s (0),y s (0),ψ s (0),u s (0),v s (0),r s (0),x a (0),y a (0),z a (0),ψ a (0),φa (0),θ a (0),u x (0),u y (0),u z (0),p a (0),q a (0), r a(0)]

[0217] =[-5m,605m,0deg,0m / s,0m / s,0rad / s,165m,705m,105m,0deg,0deg,0deg,0m / s,0m / s,0m / s,0rad / s,0rad / s,0rad / s]

[0218] Figure 4-Figure 9 They are respectively showing the simulation results of ship-UAV collaborative search under level 4 sea conditions simulated on the MATLAB simulation platform:

[0219] Figure 4 The path tracking trajectory curve of the ship-UAV collaborative circumnavigation surveillance mission is shown. Figure 4 (a) is a three-dimensional trajectory diagram, (b) is a plane trajectory diagram, Figure 4 It can be seen that the reference path of the ship-UAV is obtained according to the surround virtual guidance algorithm, which can realize the UAV to move at a constant angular speed within a fixed radius around the ship throughout the whole process. In addition, compared with a single automatic system of a ship or a UAV, since the reference signal of the UAV is generated in real time according to the position of the ship, the present invention can realize the ship-UAV collaborative tracking of the reference signal at the desired speed.

[0220] Figure 5 and Figure 6 It represents the control input of the ship-UAV collaborative system, taking into account the actuator gain of the ship-UAV system, and all control commands are accurate to the specific actuator commands.

[0221] Figure 7 , Figure 8 and Fig. 9It is the position error and attitude error of the ship-UAV collaborative system. It can be found from the figure that under the preset performance control, all system errors can eventually converge within the expected minimum value range. The position error of the UAV will produce a periodic change. This is because the circling warning task requires the UAV to follow the ship while circling the ship at a uniform angular speed. This causes the expected speed of the UAV to accelerate in the same direction as the ship and decelerate in the opposite direction in each circling cycle, thereby causing the UAV speed to change periodically. There is an unavoidable lag time from the guidance signal to the actual control of the UAV, which will eventually cause the UAV to have a periodic position error. However, under the proposed control algorithm, the final error can still converge within the expected range. Combined with the prior art, controller design and simulation experiments, the present invention has the following two beneficial effects in the field of ship-UAV collaborative circling warning:

[0222] 1) The present invention improves the existing fixed-position following alert mode of the ship-UAV system. The proposed surround virtual guidance strategy will be able to generate a ship reference signal according to the ship waypoints set in the mission, and then further autonomously plan the UAV path and ensure that the UAV always maintains a fixed distance range with the ship to meet the surround alert mission requirements. The speed of the ship and the UAV are related, which can ensure the collaborative mission of the ship-UAV. The present invention can combine the neural network system and the minimum learning parameter strategy to deal with the structural uncertainties and external environmental interference in the ship-UAV collaborative system, and at the same time design a controller for the ship-UAV system to reduce the adaptive parameters in the controller, which is beneficial to the maneuverability of the heterogeneous system. The present invention can improve the efficiency of UAV collaborative alerting of ships.

[0223] 2) The control law designed by the present invention combines preset performance control and coupled adaptive gain to achieve high-precision control under the conditions of system nonlinear terms of unknown model parameters, uncertain marine environmental interference and unknown actuator gains. The effectiveness of the guidance strategy and control algorithm proposed in the present invention was verified by conducting a simulation experiment of ship-UAV air-sea collaborative search in a simulated marine environment. The simulation results show that the algorithm has excellent performance in both control accuracy and control response speed.

[0224] 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 method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert, characterized in that: include: S1: Construct a mixed-order ship-UAV nonlinear system model as the controlled object of the control law designed in the subsequent steps; S2: Setting a waypoint path, using a virtual ship to plan a reference path according to the waypoint path, and generating a ship reference signal; obtaining the real-time position of the ship and the circling radius of the drone, and generating a drone reference signal in combination with the virtual drone; the ship reference signal includes a ship position reference signal and a ship attitude reference signal; S3: Based on the ship reference signal and the UAV reference signal, define the ship-UAV kinematic error, and design the corresponding virtual control law to eliminate the kinematic error of the ship-UAV nonlinear system model; the kinematic error includes the position error and attitude error of the ship-UAV; S4: Introducing dynamic surface technology, obtaining the dynamic surface signal of the virtual control law, and defining a dynamic error according to the dynamic surface signal; the dynamic error includes a speed error and an angular velocity error of the ship-UAV; S5: using radial basis function neural network and minimum learning parameter technology to simultaneously approximate the nonlinear terms of the ship-UAV nonlinear system model, and introducing MLP technology to simplify external interference; S6: Using the approximated simplified ship-UAV nonlinear system model and the dynamic error, design a preset performance control law and an adaptive law of actuator gain; S7: Path tracking control of the UAV in the surround warning mission is realized according to the preset performance control law and the adaptive law of the actuator gain.

2. The method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert according to claim 1 is characterized in that: A mixed-order ship-UAV nonlinear system model is constructed, as shown in formulas (1)-(5). In formula (1), H1=[x s ,y s ,ψ s ] T It represents the ship's forward movement, drift displacement and bow rolling angle, H2 = [x a ,y a ,z a ] T Indicates the forward, drift, and heave displacement of the drone, H3=[φ a ,θ a ,ψ a ] T Indicates the roll, pitch and yaw angles of the drone; I1 = [u s ,v s ,r s ] T Indicates the forward, drifting and pitching speed of the ship, I2=[u x ,u y ,u z ] T and I3=[p a ,q a ,r a ] T represents the speed and angular velocity of the drone along the ox, oz, oz axis in the coordinate system with the fuselage as the origin; F1 = [f u ,f v ,f r ] T , F2=[f x ,f y ,f z ] T and F3 = [f φ ,f θ ,f ψ ] T It represents the nonlinear term of the ship-UAV system, and its specific expansion is shown in formula (3); J1 in formula (2) is an intermediate variable; M1=diag[m u ,m v ,m r ] represents the additional mass of the ship in the forward, drifting and bow rolling directions, M2 = diag[m a ,m a ,m a ] represents the mass of the drone, M3=diag[I xx ,I yy ,I zz ] represents the rotational inertia of the drone along the ox, oy, oz axis; C1=diag[C u ,0,C r ],C2=diag[k t ,k t ,k t ],C3=diag[k t ,k t ,k d ], are all unknown system actuator gain matrices; among them, C u ,C r is the ship actuator gain, k t ,k d is the UAV actuator gain; T1=[n|n|,0,δ] T represents the ship control input, n and δ represent the ship propeller speed and rudder angle respectively, T2 and T3 are the UAV control inputs, which are specifically expanded as shown in formulas (4) and (5), where Ω is the UAV propeller rotor speed; D1=[d wu ,d wv ,d wr ] T , D2=[d wx ,d wy ,d wz ] T and D3=[d wφ ,d wθ ,d wψ ] T Respectively represent the external interference force / torque on the ship in the forward, drifting, and pitching directions and the UAV in the forward, drifting, heave, roll, pitch, and pitching directions; G = [0, 0, g] T represents the acceleration due to gravity; In formula (3), R2 = diag[R x ,R y ,R z ], R3 = diag[l,l,1], both represent the fixed actuator gain matrix, where l is the distance from the center of the drone to each rotor; d ii1 ,d ii2 ,d ii3 ,ii=u,v,r represents the nonlinear damping term of the model; k dx ,k dy ,k dz It represents the rotational drag coefficient of the drone along the ox, oy, oz axis; f u , f v , f r They represent the nonlinear terms of the ship in the forward, drift and bow rolling directions, respectively, and f x , f y , f z They represent the nonlinear terms of the UAV in the forward, drift, and heave directions, respectively, and f φ ,f θ ,f ψ They represent the nonlinear terms of the UAV in the roll, pitch and yaw directions, R x ,R y ,R z Respectively represent the gain coefficients of the drone on ox, oy, and oz.

3. The method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert according to claim 2 is characterized in that: Setting a waypoint path, using a virtual ship to plan a reference path according to the waypoint path, and generating a ship reference signal; Obtain the real-time position of the ship and the UAV's circling radius, and generate a UAV reference signal in combination with the virtual UAV, including: S21, set the waypoint path and the ship turning radius, and select the target point P according to the set turning radius before and after each waypoint except the starting point and the end point inwk and P outWk ; S22, using a virtual ship to plan a reference path according to the waypoint path, as shown in formula (6), In the formula, x sd ,y sd ,ψ sd Indicates the forward, drifting distance and heading angle of the virtual ship L1, that is, the ship position reference signal, u sd ,r sd represents the forward speed and yaw angular velocity of L1 ship, u sd According to the actual situation, set sd The calculation formula is shown in formula (7). Where, L 1s is the distance from L1 ship to the target point, η s =ψ sd -ψ L1s , ψ L1s is the azimuth from L1 ship to the target point, η s Indicates the deviation angle between L1 ship and target point; a ls represents the lateral acceleration of the L1 ship; The azimuth from the actual ship to the L1 ship is shown in formula (8): In the formula, ψ r Indicates the azimuth from the actual ship to the L1 ship, i.e., the ship attitude reference signal, x se =x sd -x s ,y se =y sd -y s , indicating the forward and drift distance error of the ship; S23, obtaining the real-time position of the ship and the circling radius of the drone; S24, using the virtual drone to obtain the drone's reference signal, as shown in formula (9), In the formula, x ad ,y ad ,ψ ad represents the forward and drift distances and heading angles of the virtual UAV, i.e., the reference signal of the UAV, T a is the set UAV orbit period, L R is the set UAV orbit radius, η a is the azimuth from the ship to the virtual drone, pp(0) its initial value, x ae =x ad -x a ,y ae =y ad -y a , is the UAV's forward and lateral distance error, z ad is the desired rise and fall altitude of the drone.

4. The method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert according to claim 3 is characterized in that: According to the ship reference signal and the UAV reference signal, the ship-UAV kinematic error is defined, and the corresponding virtual control law is designed, including: S31. Define the ship-UAV position error based on the ship reference signal and the UAV reference signal, as shown in formula (10): In the formula, z se represents the ship position error, z ae Indicates the heave distance error of the drone; S32, deriving the position error of the ship-UAV, as shown in formula (11), In the formula, ψ se =ψ r -ψ s is the ship’s heading angle error; S33, defining a preset performance function of the ship-UAV position error, as shown in formula (12), In the formula, ρ i ,i=u,x,y,z, is the preset performance function of position error, ρ i0 and ρ i∞ is i The initial value and final convergence value of k αi is a constant; Define the conversion error of the ship-UAV position error as shown in formula (13): In the formula, σ i ,i=u,x,y,z represents the conversion error of position error; S34. Design the virtual control law of the position error of the ship-UAV, as shown in formula (14): In the formula, k αu ,k αx ,k αy ,k αz is the virtual control law α u ,α x ,α y ,α z design parameters.

5. The method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert according to claim 4 is characterized in that: Define the kinematic errors of the ship-UAV and design the corresponding virtual control law, including: S35. Define the ship-UAV attitude error according to the ship reference signal and the UAV reference signal, as shown in formula (15): In the formula, ψ se Indicates the ship's heading angle error, φ ae represents the roll angle error of the UAV, θ ae represents the pitch angle error of the UAV, ψ ae represents the UAV heading angle error; ψ ad represents the reference roll angle of the UAV, θ ad Indicates the reference pitch angle of the drone; S36, deriving the attitude error of the ship-UAV, as shown in formula (16), S37, defining a preset performance function of the ship-UAV attitude error, as shown in formula (17), r j (t)=(ρ j0 -r j∞ )e -kjt +r j∞ (17) In the formula, ρ j ,j=r,φ,θ,ψ, is the preset performance function of attitude error, ρ j0 and ρ j∞ is j The initial value and final convergence value of k αj is a constant; Define the conversion error of the ship-UAV attitude error as shown in formula (18): In the formula, σ j ,j=r,φ,θ,ψ, represents the attitude error conversion error; S38. Design the virtual control law of the ship-UAV attitude error, as shown in formula (19): In the formula, k αr ,k αφ ,k αθ ,k αψ is the virtual control law of attitude error α r ,α φ ,α θ ,α ψ control parameters.

6. The method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert according to claim 5 is characterized in that: Introducing dynamic surface technology, obtaining the dynamic surface signal of the virtual control law, and defining the dynamic error according to the dynamic surface signal, including: S41. Introduce dynamic surface technology, reduce the derivative of the virtual control law of the position error and the derivative of the virtual control law of the attitude error, and obtain the dynamic surface error of the position error and the dynamic surface error of the attitude error, as shown in formulas (20) and (21). In formula (20), β i is α i The dynamic surface signal, ∈ i is the time constant, and the dynamic surface error of the position error is y i =α i -β i , α i A virtual control law representing the position error; In formula (21), β j is α j The dynamic surface signal, ∈ j is the time constant, and the dynamic surface error of the attitude error is y j =α j -β j , α j A virtual control law representing the attitude error; S42. Define the ship-UAV velocity error and take its derivative, as shown in formulas (22) and (23): In the formula, u se ,u xe ,u ye ,u ze represents the ship's forward movement, the UAV's forward movement, drift and heave speed errors, B2 = [β x ,β y ,β z ] T Dynamic surface signal matrix representing position error; Define the angular velocity error between the ship and the UAV and take its derivative, as shown in equations (24) and (25): In the formula, r se ,p ae ,q ae ,r ae Indicates the angular velocity error of the ship's bow roll, the UAV's roll, pitch and bow roll; B3 = [β φ ,β θ ,β ψ ] T , represents the dynamic surface signal matrix of the attitude error.

7. The method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert according to claim 6 is characterized in that: The nonlinear terms of the ship-UAV nonlinear system model are approximated by radial basis function neural network and minimum learning parameter technology, and the MLP technology is introduced to simplify the external interference, as shown in formulas (26) and (27): f i -d wi ≤Θ i1 Θ i2 -c i S i W i (26) f j -d wj ≤Θ j1 Θ j2 -c j S j W j (27) In the formula, f i and f j represents the nonlinear term of the ship-UAV nonlinear system model, Θ i2 =1+‖S i ‖+1,Θ j2 =1+‖S j ‖+1,Θ i1 ,Θ j1 ,Θ i2 ,Θ j2 It is the intermediate variable in the minimum learning parameter technique; A i represents f i The neural network weight update law, A j represents f j The neural network weight update law, c i ,c j A i ,A j The norm of S i and S j represents the Gaussian function, ε i and ε j represents the approximation error, is the approximation error ε i The maximum value of is the approximation error ε j The maximum value, d mi It is the external environment interference wi The maximum value, d mj It is the external environment interference wj The maximum value of W i ,W j represents the intermediate variable, as shown in formula (28), 8. The method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert according to claim 1 is characterized in that: Using the approximated simplified ship-UAV nonlinear system model and the dynamic error, a preset performance control law and an adaptive law of actuator gain are designed, including: S61, obtaining the actual control input, and converting the actual control input into a preset performance control law and an adaptive law of actuator gain, as shown in formula (29), In the formula, τ l ,l=u,r,x,y,z,φ,θ,ψ are the preset performance control laws of the ship's forward direction, bow angle, UAV's forward direction, drift, heave direction, roll, pitch and bow angle; n represents the propeller speed of the ship; δ represents the rudder angle of the ship; λ2=[λ kt ,λ kt ,λ kt ] T ,λ3=[λ kt ,λ kt ,λ kd ] T , They are λ cu ,λ cr ,λ kt ,λ kd The estimated value of cu ,λ cr ,λ kt ,λ kd They represent the adaptive parameters of the actuator gain, i.e., the adaptive law, and represents the matrix of actuator gains, τ2 and τ3 represent the preset performance control law matrices; S62. Using the simplified ship-UAV nonlinear system model, formulas (23) and (29), the MLP technology, coupling gain adaptation technology and backstepping method are introduced to design the preset performance control law of the speed error and the adaptive law of the actuator gain, as shown in formulas (30)-(32). In the formula, k i ,Γ Cu1 ,Γ Cu2 ,θ i1 ,θ i2 is the design parameter, ω i represents the adaptive parameter introduced by the minimum learning parameter, Yes i An estimated value of Ψ i Represents ω i The robust damping term of i1 ,ρ i2 is a constant; τ u ,τ x ,τ y ,τ z represents the preset performance control law for the speed error, represents the derivative of the adaptive parameter; S63. Using the simplified ship-UAV nonlinear system model, formulas (25) and (29), the MLP technology, coupling gain adaptation technology and backstepping method are introduced to design the preset performance control law of the angular velocity error and the adaptive law of the actuator gain, as shown in formulas (33)-(35). In the formula, k j ,Γ Cr1 ,Γ Cr2 ,Γ kt1 ,Γ kt2 ,Γ kd1 ,Γ kd2 ,θ j1 ,θ j2 is the design parameter, Yes j The estimated value of ω j The robust damping term of j1 ,ρ j2 is a constant; τ j Denotes the preset performance control law for the angular velocity error, ω j Adaptive parameters introduced for the minimum learning parameter technique, The derivative of the adaptive parameter representing the actuator gain.

9. The method for ship-air-sea collaborative mission guidance and gain adaptive control for surround alert according to claim 8, characterized in that: The nonlinear decoupling technology is used to solve formula (30) to obtain the reference roll angle and pitch angle of the UAV, as shown in formula (36): In the formula, φ ad represents the reference roll angle of the UAV, θ ad Indicates the reference pitch angle of the drone.

Images