Ship-based unmanned aerial vehicle wake suppression method based on unknown input reconstruction compensation

Abstract

The application discloses a ship-borne unmanned aerial vehicle ship wake restraining method based on unknown input reconstruction compensation, and steps are as follows: S1, selecting different trimming points, trimming a longitudinal channel of a nonlinear model of the ship-borne unmanned aerial vehicle to obtain a linear switching system model; S2, designing a switching stabilization control law for the linear switching system to make the system asymptotically stable; S3, modeling external disturbances such as ship wakes in a landing stage as unknown inputs and introducing the unknown inputs into the switching system, and designing a state observer for the system; S4, based on state observation results, reconstructing the unknown inputs by adopting a super-spiral sliding mode method; and S5, based on the reconstructed unknown inputs, introducing unknown input compensation in the switching stabilization control law to restrain influences of the disturbances such as ship wakes. The application can reconstruct and compensate influences of unknown disturbances such as ship wakes on the ship-borne unmanned aerial vehicle in a landing stage, and realizes smooth and safe landing of the ship-borne unmanned aerial vehicle.

Description

Technical Field

[0001] This invention relates to the field of flight control technology, and in particular to a method for suppressing the wake of a shipborne unmanned aerial vehicle based on unknown input reconstruction compensation. Background Technology

[0002] The landing process for carrier-based aircraft is significantly more complex than that for land-based aircraft. Firstly, an aircraft carrier is a six-degree-of-freedom platform at sea, experiencing planar motion on the sea surface and rolling, pitching, and heave movements due to waves. Secondly, the wake field, also known as the stern flow, is an airflow field generated above the stern of the angled deck of the aircraft carrier through the interaction of atmospheric and hull motion. Its effective range is generally one kilometer from the stern of the deck, and its density variation is determined by the hull shape and the direction of airflow. Furthermore, aircraft carriers typically employ an asymmetrical design, with the island superstructure usually located on the starboard side and the deck on the port side. This asymmetrical design leads to an asymmetrical wake, causing different airflow effects on the wings during approach and landing, increasing the difficulty of approach and landing control. Complex sea conditions and adverse environments cause severe deck motion and wake flow, causing the ideal landing point to drift, altering the glide slope command, and creating external interference with aircraft speed tracking, severely impacting the accuracy of carrier-based aircraft landings.

[0003] Shipborne unmanned aerial vehicles (UAVs) are subject to various disturbances during their final landing maneuvers. Suppressing the impact of wake disturbances is crucial. The technical challenges lie in the fact that the wake field is influenced by numerous and constantly changing factors, exhibiting randomness and strong nonlinearity. The direction and magnitude of the wake are complex, making modeling relatively difficult. Currently, research on wake suppression is limited, and practical methods for wake estimation and suppression are still lacking. Therefore, accurate estimation and suppression of wake are critical issues that urgently need to be addressed. Summary of the Invention

[0004] Purpose of the invention: The purpose of this invention is to provide a method for suppressing the wake of shipborne UAVs based on unknown input reconstruction compensation, which can accurately estimate disturbance signals such as the wake and make the landing of shipborne UAVs smooth and safe.

[0005] Technical solution: The wake suppression method for shipborne unmanned aerial vehicles of the present invention includes the following steps:

[0006] S1. Select different balancing points to balance the longitudinal channel of the nonlinear model of the shipborne UAV and obtain the linear switching system model.

[0007] S2, a switching stabilization control law is designed for linear switching systems to stabilize the system;

[0008] S3 models the external disturbance of the ship's wake during the landing phase as an unknown input and introduces it into the linear switching system, and designs a state observer for the linear switching system.

[0009] S4. Based on the state observation results, the unknown input is reconstructed using the super-spiral sliding mode method;

[0010] S5 introduces unknown input compensation in the switching stabilization control law based on the reconstructed unknown input to suppress the influence of wake disturbance.

[0011] Furthermore, in step S1, the longitudinal nonlinear dynamic model of the shipborne UAV landing section is expressed as follows:

[0012]

[0013] Wherein, [V,α,q,θ,H] T These are the five states of the system; the superscript T represents the transpose operation; V is the airspeed of the carrier-based UAV, α is the angle of attack, q is the pitch rate, θ is the pitch angle, and H is the flight altitude; δ e Let f be the elevator deflection angle, F be the engine thrust, and f() be a nonlinear function satisfying the dynamics and kinematics of the shipborne UAV.

[0014] Based on the different landing altitudes, the nonlinear system is divided into segments, resulting in a longitudinal nonlinear switching model. For the nonlinear switching system, flight altitudes H of 75m, 45m, and 15m are selected for trimming, yielding a linear switching system model for the carrier-based UAV.

[0015]

[0016] y(t)=C σ(t) x(t)

[0017] Among them, x(t)=[ΔV,Δα,Δq,Δθ] T These are the four states of a linearly switched system, where ΔV, Δα, Δq, and Δθ are the increments of airspeed V, angle of attack α, pitch angular velocity q, and pitch angle θ from the corresponding state at the trim point, respectively; u(t) = [Δδ e ,ΔF] T Δδ is the control input of the system. e ΔF and ΔF are the elevator deflection angles δ and δ, respectively. e The increment of the control input corresponding to the deviation of the engine thrust F from the trim point; w(t) is the external disturbance from the wake of the ship that the system experiences, which is an unknown input; A σ(t) B σ(t) C σ(t) and D σ(t)These are the constant matrices obtained after linearization near the balancing point; y(t) is the system output; σ(t) is the piecewise constant discrete state function in the discrete set. The value N is the number of switching system subsystems.

[0018] Furthermore, in step S2, it is assumed that the unknown input w(t) in the linear switching system model of the shipborne UAV is ≡ 0, that is, w(t) is always 0. At this time, the state equation degenerates into:

[0019]

[0020] The switching stabilization control law is designed as follows:

[0021] u(t) = K σ(t) x(t)

[0022] K σ(t) It is the control gain to be designed.

[0023] Furthermore, in step S3, the external disturbance of the ship's wake during the landing phase is modeled as an unknown input and introduced into the linear switching system. A state observer is designed for this linear switching system, and the design steps are as follows:

[0024] First, a coordinate transformation of the state and output is introduced to decouple the unknown input w(t) from the transformed subset of state coordinates; then, a transformation matrix T is introduced. q and U q , The definition is as follows:

[0025]

[0026] Where, for matrix Γ, Γ ⊥ Indicates that Γ is satisfied ⊥ A matrix where Γ = 0, Γ + Denote the left pseudo-inverse of Γ, satisfying Γ + =(Γ T Γ) -1 Γ T ;

[0027] Matrix T q A non-singular square matrix has an inverse matrix:

[0028]

[0029] Let q = σ(t), and introduce a non-singular linear transformation. and These are the transformed state vector and output vector, which are divided into the following:

[0030]

[0031]

[0032] Therefore, the transformed system is:

[0033]

[0034]

[0035]

[0036]

[0037] Wherein, the block matrix A σ(t),1 A σ(t),2 A σ(t),3 A σ(t),4 B σ(t),1 B σ(t),2 , satisfy:

[0038]

[0039] state Through output Directly obtained, reconstructed Obtain the transformed state Furthermore, all information about x(t) can be obtained through the following inverse transformation.

[0040]

[0041] In order to estimate The information is defined as the observed values ​​of state x(t). The observer is designed as follows:

[0042]

[0043] in, These are the observations of the system's four states. For the airspeed observation value of the shipborne UAV, These are the observed angles of attack. These are the observed values ​​of pitch angular velocity. These are the observed values ​​for the pitch angle; Defined as:

[0044]

[0045] in, yes The observed values, matrix L σ(t) The observer gain to be designed; the observer error is defined. It satisfies:

[0046]

[0047] Furthermore, the observer gain L σ(t) Solve using the following steps:

[0048]

[0049]

[0050] Among them, P q It is a Lyapunov matrix, Y q It is the intermediate variable matrix; further, the observer gain μ > 1 represents the growth coefficient of the multi-Lyapunov function at the switching moment; α > 0 represents the decay rate of the multi-Lyapunov function during the operation of the subsystem; then the obtained observer gain matrix L q This can reduce errors The observed values ​​of the state converge exponentially. With the exponent approaching the state

[0051] Furthermore, in step S4, based on the state observation results, the unknown input is reconstructed using the superspiral sliding mode method. The design steps are as follows:

[0052] Based on the observed values ​​of the state Reconstructing unknown inputs based on the superspiral sliding mode method, defining variables satisfy

[0053]

[0054] Among them, v w (t) is the input variable to be designed;

[0055] Based on this, define the sliding mode function s(t):

[0056]

[0057] After a limited time... Combined with state equations The derivative of s(t) can be derived as follows:

[0058]

[0059] v w (t) can be designed in the following form:

[0060]

[0061]

[0062] Wherein, the constant w dmax Need to meet λ1 and λ2 are constant values ​​greater than 0. By choosing appropriate λ1 and λ2, it is guaranteed that there exists a finite time t. d Make the following equations true:

[0063]

[0064] Therefore, according to the above formula, accurate reconstruction of unknown input is achieved within a finite amount of time, that is,

[0065]

[0066] Based on the above conditions, the solution is obtained. Achieve reconstruction of unknown inputs.

[0067] Furthermore, in step S5, based on the reconstructed unknown input, unknown input compensation is introduced into the switching stabilization control law to suppress the influence of wake disturbance; the design steps are as follows:

[0068] Based on the reconstructed unknown input, an additional unknown input compensation term is introduced in the switching stabilization control law:

[0069]

[0070] Applying the above equation to the system, the state equation of the closed-loop system is:

[0071]

[0072] Therefore, the external disturbances in the wake of the ship were effectively suppressed by the method of reconstruction compensation through unknown input.

[0073] Compared with the prior art, the significant advantages of this invention are as follows:

[0074] 1. This invention models the external disturbances such as the wake turbulence experienced by carrier-based aircraft during landing as unknown inputs, and further designs a switching state observer. Based on this, the super-spiral sliding mode method is applied to reconstruct the unknown inputs, which can estimate the disturbance signals such as the wake turbulence more accurately.

[0075] 2. By introducing additional unknown input compensation into the controller, compared with the traditional control scheme without introducing unknown input compensation, the control method proposed in this invention can effectively suppress the influence of external disturbances such as ship wake, and has better tracking accuracy, making the landing of shipborne UAVs more stable and safe. Attached Figure Description

[0076] Figure 1This is a diagram showing the overall structure of a shipborne UAV wake suppression control scheme based on unknown input reconstruction compensation.

[0077] Figure 2 A comparison diagram showing the first component of the unknown input reconstructed using this invention and the first component of the actual disturbance;

[0078] Figure 3 A comparison diagram of the second component of the unknown input reconstructed using the present invention and the second component of the actual disturbance;

[0079] Figure 4 The image shows the airspeed response curve of the shipborne UAV during the landing phase obtained using this invention.

[0080] Figure 5 Angle of attack response curve of shipborne UAV during landing phase obtained using the present invention;

[0081] Figure 6 The diagram shows the pitch angular velocity response curve of the shipborne UAV during the landing phase obtained using this invention.

[0082] Figure 7 The diagram shows the pitch angle response curve of the shipborne UAV during the landing phase obtained using this invention.

[0083] Figure 8 The image shows the altitude response curve of the shipborne UAV during the landing phase obtained using this invention. Detailed Implementation

[0084] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific examples. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. The invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0085] For complex nonlinear systems, a common design approach is to linearize the nonlinear model by selecting different trim points based on the operating point requirements. However, during the landing phase, the aerodynamic data of carrier-based UAVs are constantly changing, making it difficult for a single model to accurately describe the system's flight dynamics. Using multiple subsystems of a switching system can effectively solve this problem. Secondly, external disturbances such as the wake turbulence during landing are significant factors affecting safe landing. By designing an observer to reconstruct and compensate for these disturbances, the risks during landing can be effectively reduced. Therefore, this invention models the longitudinal kinematics and dynamics of the carrier-based UAV as a linear switching system and models external disturbances such as the wake turbulence as unknown inputs. The unknown inputs are reconstructed based on the super-spiral sliding mode method to compensate for the influence of external disturbances such as the wake turbulence. The overall structural diagram of this invention is shown below. Figure 1 As shown.

[0086] The wake suppression method for shipborne unmanned aerial vehicles based on unknown input reconstruction compensation of the present invention specifically includes the following steps:

[0087] Step 1, the longitudinal nonlinear dynamic model of the carrier-based UAV landing section, is expressed as follows:

[0088]

[0089] Wherein, [V,α,q,θ,H] T These are the five states of the system: T represents the transpose operation; V is the airspeed of the carrier-based UAV, α is the angle of attack, q is the pitch rate, θ is the pitch angle, and H is the flight altitude; δ e Let f be the elevator deflection angle, F be the engine thrust, and f() be a nonlinear function satisfying the dynamics and kinematics of the shipborne UAV.

[0090] Based on the different landing altitudes, the nonlinear system is divided into segments, resulting in a longitudinal nonlinear switching model. For the nonlinear switching system, flight altitudes H of 75m, 45m, and 15m are selected for trimming, yielding a linear switching system model for the carrier-based UAV.

[0091]

[0092] In the nonlinear dynamic model, the state variable height H can be calculated from the velocity V, angle of attack α, and pitch angle θ. Therefore, the above-mentioned fifth-order nonlinear dynamic model can be described by a fourth-order linear switching system model, where x(t)=[ΔV,Δα,Δq,Δθ]. T These are the four states of a linearly switched system, where ΔV, Δα, Δq, and Δθ are the increments of airspeed V, angle of attack α, pitch angular velocity q, and pitch angle θ from the corresponding state at the trim point, respectively; u(t) = [Δδ e ,ΔF] T Δδ is the control input of the system. e ΔF and ΔF are the elevator deflection angles δ and δ, respectively. e The increment of the control input corresponding to the deviation of the engine thrust F from the trim point; w(t) is the external disturbance such as the wake turbulence experienced by the system, which is regarded as an unknown input; A σ(t) B σ(t) C σ(t) and D σ(t) These are the constant matrices obtained after linearization near the balancing point; y(t) is the system output; σ(t) is the piecewise constant discrete state function, which is defined on the discrete set. The value N is the number of switching system subsystems.

[0093] Step 2, assuming the unknown input w(t) in the linear switching system model of the shipborne UAV is ≡ 0, that is, w(t) is always 0, the state equation (i.e., formula (2) for the linear switching system model of the shipborne UAV) degenerates into:

[0094]

[0095] The switching stabilization control law is designed as follows:

[0096] u(t) = K σ(t) x(t) (4)

[0097] K σ(t) It is the control gain to be designed, which can be obtained through the stabilization control design method of a general linear switching system.

[0098] Step 3: Model external disturbances such as the wake during the landing phase as unknown inputs and introduce them into the linear switching system. Design a state observer for this linear switching system. The design steps are as follows:

[0099] First, a coordinate transformation of the state and output is introduced to decouple the unknown input w(t) from the transformed subset of state coordinates, and a transformation matrix T is introduced. q and U q , The definition is as follows:

[0100]

[0101] Where, for matrix Γ, Γ ⊥ Indicates that Γ is satisfied ⊥ A matrix where Γ = 0, Γ + Denote the left pseudo-inverse of Γ, satisfying Γ + =(Γ T Γ) -1 Γ T ;

[0102] Matrix T q A non-singular square matrix has an inverse matrix:

[0103]

[0104] Let q = σ(t), and introduce a non-singular linear transformation. and These are the transformed state vector and output vector, which are divided into the following:

[0105]

[0106]

[0107] Therefore, the transformed system is:

[0108]

[0109] Wherein, the block matrix A σ(t),1 A σ(t),2 A σ(t),3 A σ(t),4 B σ(t),1 B σ(t),2 , satisfy:

[0110]

[0111] state It can be output Directly obtained, reconstructed The transformed state can then be obtained. Furthermore, all information about x(t) can be obtained through the following inverse transformation.

[0112]

[0113] In order to estimate The information is defined as the observed values ​​of state x(t). The observer is designed as follows:

[0114]

[0115] in, These are the observations of the system's four states. For the airspeed observation value of the shipborne UAV, These are the observed angles of attack. These are the observed values ​​of pitch angular velocity. These are the observed values ​​for the pitch angle; Defined as:

[0116]

[0117] in, yes The observed values, matrix L σ(t) The observer gain to be designed; the observer error is defined. It satisfies:

[0118]

[0119] Furthermore, the observer gain L q (i.e. L) σ(t) ), The solution can be found by following these steps:

[0120]

[0121]

[0122] Among them, P q It is a Lyapunov matrix, Y q These are intermediate variable matrices, which can be obtained by solving the system of linear matrix inequalities in equations (15) and (16). Furthermore, the observer gain... μ q,r >1 indicates the growth coefficient of the multi-Lyapunov function at the switching moment; α>0 indicates the decay rate of the multi-Lyapunov function during subsystem operation; the observer gain matrix L obtained by the above method q This can reduce errors Exponential convergence, i.e., the observed values ​​of the state With the exponent approaching the state

[0123] Step 4: Based on the state observation results, the unknown input is reconstructed using the superspiral sliding mode method. The design steps are as follows:

[0124] Based on the observed values ​​of the state Reconstructing unknown inputs based on the superspiral sliding mode method, defining variables satisfy

[0125]

[0126] Among them, v w (t) is the input variable to be designed;

[0127] Based on this, define the sliding mode function s(t):

[0128]

[0129] After a limited time... Combined with state equations The derivative of s(t) can be derived as follows:

[0130]

[0131] v w (t) can be designed in the following form:

[0132]

[0133] Wherein, the constant w dmax Need to meet `sign()` is the sign function; `λ1` and `λ2` are constant values ​​greater than 0. By choosing appropriate `λ1` and `λ2`, it can be guaranteed that there exists a finite time `t`. dMake the following equations true:

[0134]

[0135] Therefore, according to the equation in formula (21), accurate reconstruction of unknown input is achieved within a finite amount of time, that is,

[0136]

[0137] According to the conditions of formula (22), the following can be calculated: This allows for the reconstruction of unknown inputs.

[0138] Step 5: Based on the reconstructed unknown input, introduce unknown input compensation into the switching stabilization control law to suppress the effects of disturbances such as the wake turbulence. The design steps are as follows:

[0139] Based on the reconstructed unknown input, an additional unknown input compensation term is introduced in the switching stabilization control law:

[0140]

[0141] Applying equation (23) to a linearly switched system, the state equation of the closed-loop linearly switched system is:

[0142]

[0143] It can be seen that external disturbances such as the wake turbulence are effectively suppressed through the unknown input reconstruction compensation method, thereby ensuring the safe and stable landing of shipborne UAVs.

[0144] To verify the effectiveness of this invention in flight control during the landing phase of shipborne UAVs, the following simulation experiment was conducted.

[0145] In this embodiment, the disturbance variable w(t) is first randomly generated by simulating the wake pattern of the ship. Here, we consider the disturbance variable w(t) = [w1(t)w2(t)]. T It is a vector containing two perturbation components. Based on this invention, the perturbation components are reconstructed and compared with the randomly generated unknown perturbation components, for example... Figure 2 , Figure 3 As shown, the disturbance reconstruction method designed in this invention can effectively estimate unknown inputs such as the wake.

[0146] This embodiment simulates and tests the glide performance of the altitude tracking system under external disturbances. Ignoring the ship's deck motion, a trim state at an angle of attack of 8.39° is selected, with an initial pitch angle of 4.87°. The simulation time is the last 20 seconds of the landing phase. Stern airflow is introduced approximately 12.5 seconds before landing, at a distance of 800m from the ship. The comparison of the uncompensated and compensated state response curves is shown in the figure below. Figure 4 ,Figure 5 , Figure 6 , Figure 7 , Figure 8 As shown.

[0147] It can be seen that without suppressing external disturbances such as the wake turbulence, the various state variables of the shipborne UAV exhibit significant chattering. The unknown input compensation method designed in this invention can effectively suppress the chattering of various state variables and mitigate the adverse effects caused by external disturbances such as the wake turbulence.

[0148] according to Figure 4 medium speed Figure 5 medium angle of attack and Figure 7 The mid-pitch angle response allows for the calculation of the altitude response curve. It can be seen that at the end of the landing phase, without suppressing external disturbances such as the wake turbulence, the carrier-based UAV deviates significantly from the desired uniform descent trajectory, which is detrimental to safe landing. By introducing the wake turbulence compensation method designed in this invention, the glide path trajectory is significantly corrected, exhibiting the desired uniform descent characteristics. Therefore, the UAV can complete the landing mission more safely.

Claims

1. A method for suppressing the wake of a shipborne unmanned aerial vehicle (UAV) based on unknown input reconstruction compensation, characterized in that, Includes the following steps: S1. Select different balancing points to balance the longitudinal channel of the nonlinear model of the shipborne UAV and obtain the linear switching system model. S2, a switching stabilization control law is designed for linear switching systems to make the system asymptotically stable; S3 models the external disturbance of the ship's wake during the landing phase as an unknown input and introduces it into the linear switching system, and designs a state observer for the linear switching system. S4. Based on the state observation results, the unknown input is reconstructed using the super-spiral sliding mode method; S5, based on the reconstructed unknown input, introduces unknown input compensation in the switching stabilization control law to suppress the influence of wake disturbance; In step S1, the longitudinal nonlinear dynamic model of the landing section of the shipborne UAV is expressed as follows: ， in, These are the 5 states of the system. This represents the transpose operation; For the airspeed of the shipborne drone, For the angle of attack, The pitch angular velocity, The pitch angle, Flight altitude; For elevator deflection angle, For engine thrust; To satisfy the nonlinear functions of shipborne UAV dynamics and kinematics; Based on the different altitudes during the landing phase, the nonlinear system is divided into segments, resulting in a longitudinal nonlinear switching model. For each nonlinear switching system, a flight altitude is selected... for , , After balancing, the linear switching system model of the shipborne UAV is obtained: ， in, These are the four states of a linearly switching system. , , and airspeed Angle of attack Pitch angular velocity and pitch angle The increment of the deviation from the corresponding state at the balance point; For the system's control input, and These are the elevator deflection angles. and engine thrust The increment of the corresponding control input at the deviation from the balance point; The external disturbance from the ship's wake that the system receives is an unknown input; , , and These are the constant matrices obtained after linearization near the balancing point; It is the system output; It is a piecewise constant discrete state function, in discrete sets Take the value above. This refers to the number of system subsystems that can be switched. In step S2, we assume the unknown input in the linear switching system model of the shipborne UAV. At this point, the state equation of the shipborne UAV linear switching system model degenerates into: ， The switching stabilization control law is designed as follows: ， It is the control gain to be designed; In step S3, the external disturbance of the ship's wake during the landing phase is modeled as an unknown input and introduced into the linear switching system. A state observer is designed for this linear switching system. The design steps are as follows: First, we introduce coordinate transformations for the state and output to transform the unknown input. Decoupled from the transformed subset of state coordinates; introduce a transformation matrix. and , The definition is as follows: ， For the matrix , Indicates satisfaction The matrix, express The left pseudo-inverse satisfies ; matrix A non-singular square matrix has an inverse matrix: ， make Introducing non-singular linear transformation , , and These are the transformed state vector and output vector, which are divided into the following: ， ， Therefore, the transformed system is: ， Wherein, the block matrix , , , , , , satisfy: ， state Through output Directly obtained, reconstructed To obtain the transformed state All the information is further obtained through the following inverse transformation. Full information: ， In order to estimate Information, defining state The observed value is The observer is designed as follows: ， in, These are the observations of the system's four states. For the airspeed observation value of the shipborne UAV, These are the observed angles of attack. These are the observed values ​​of pitch angular velocity. These are the observed values ​​for the pitch angle; Defined as: ， in, yes The observed values, matrix The observer gain to be designed; the observer error is defined. It satisfies: ， Furthermore, observer gain Solve using the following steps: ， ， in, It is a Lyapunov matrix. It is the intermediate variable matrix; further, the observer gain , ; This represents the growth coefficient of the multi-Lyapunov function at the switching moment; The decay rate of the multi-Lyapunov function during the operation of the subsystem is represented by the observer gain matrix. This can reduce errors The observed values ​​of the state converge exponentially. With the exponent approaching the state ; In step S4, based on the state observation results, the unknown input is reconstructed using the superspiral sliding mode method. The design steps are as follows: Based on the observed values ​​of the state Based on the superspiral sliding mode method, unknown inputs are reconstructed, and variables are defined. ,satisfy ， in, These are the input variables to be designed; Based on this, define the sliding mode function. : ， After a limited time... Combined with state equations , deduced The derivative is as follows: ， The design is as follows: ， Where, constant Need to meet ; sign() is the sign function; and It is a constant value greater than 0, and by selecting an appropriate... and Guarantee the existence of a finite time. Make the following equations true: ， Therefore, according to the above formula, accurate reconstruction of unknown input is achieved within a finite amount of time: ， Then the solution is calculated This enables the reconstruction of unknown inputs.

2. The method for suppressing the wake of a shipborne unmanned aerial vehicle based on unknown input reconstruction compensation according to claim 1, characterized in that, In step S5, based on the reconstructed unknown input, unknown input compensation is introduced into the switching stabilization control law to suppress the influence of wake disturbance; the design steps are as follows: Based on the reconstructed unknown input, an additional unknown input compensation term is introduced in the switching stabilization control law: ， Applying the above equation to a linearly switched system, the state equation of the closed-loop linearly switched system is: ， Therefore, the external disturbances in the wake of the ship were effectively suppressed by the method of reconstruction compensation through unknown input.

Images