Smooth switching control method of comprehensive anti-jamming transition section of tilt-rotor unmanned aerial vehicle

Abstract

The application discloses a comprehensive anti-interference smooth switching control method for a transition section of a tilt-rotor unmanned aerial vehicle, and steps are as follows: S1, a longitudinal nonlinear dynamics model of the transition section of the tilt-rotor unmanned aerial vehicle is established, different control points are selected in a tilt corridor to perform trimming, and a switching linear system model is obtained; S2, an external system is introduced to represent disturbance input generated by rotor wake; S3, a disturbance observer is designed to estimate the disturbance from the external system; S4, a reference model is designed to convert the longitudinal control problem of the transition section into a comprehensive anti-interference control problem of the switching linear system, and a comprehensive anti-interference control law is designed; and S5, a smooth interpolation control strategy is designed to suppress state chattering caused by controller switching. The application can suppress various disturbances affecting the tilt-rotor unmanned aerial vehicle in a transition process, suppress state chattering caused by controller switching, and make the transition process of the tilt-rotor unmanned aerial vehicle more stable and safe.

Description

Technical Field

[0001] This invention relates to the field of flight control technology, and in particular to a comprehensive anti-interference smooth switching control method for the transition phase of a tiltrotor unmanned aerial vehicle. Background Technology

[0002] Tiltrotor aircraft, by installing two rotating nacelles at the wingtips, can switch between helicopter and fixed-wing modes. While combining the advantages of both, tiltrotor aircraft effectively solve the problems of short range, slow speed, and high noise levels associated with helicopters, as well as the demanding takeoff and landing conditions and difficulties in complex terrain faced by fixed-wing aircraft. Research on tiltrotor aircraft has significant practical importance and promising development prospects.

[0003] Tiltrotor aircraft have three flight modes: helicopter mode, transition mode, and fixed-wing mode. In helicopter mode, the nacelle is vertical, and the tiltrotor uses the lift generated by the rotor to counteract gravity, enabling vertical takeoff and landing, as well as hovering. The process of switching from helicopter to fixed-wing mode, or vice versa, is called the transition process, which is accomplished by tilting the nacelle. In fixed-wing mode, the nacelle is horizontal, and the wings generate lift to overcome gravity, while the rotor rotation provides forward thrust, enabling high-speed flight. By switching between helicopter and fixed-wing modes, tiltrotor aircraft can achieve the functions required in specific situations; for example, switching to fixed-wing mode for high-speed flight, or switching to helicopter mode for vertical takeoff and landing, hovering, etc.

[0004] The advantages of tiltrotor aircraft enable them to cover a variety of flight missions and even extend to flight missions in various complex situations and terrains, showing broad development prospects. However, due to their complex structure, related technical research faces many challenges, the most significant being the flight dynamics modeling and flight control law design during transition modes. First, the need to simultaneously consider the dynamic characteristics of helicopters and fixed-wing aircraft, and the significant changes in aerodynamic shape caused by the nacelle tilting during transition modes, with its strong nonlinear characteristics and drastic changes in aerodynamic parameters, make modeling even more difficult. Second, during actual nacelle tilting, the aircraft is subject to various disturbances, including rotor wake, affecting flight stability. Furthermore, the complex flight control methods make designing suitable control laws to achieve smooth transitions between flight modes a major challenge. Summary of the Invention

[0005] Purpose of the invention: The purpose of this invention is to provide a control method for smooth switching of the transition section of a tiltrotor UAV with comprehensive anti-interference capabilities, enabling a smooth transition in the transition section.

[0006] Technical solution: The tilt-rotor unmanned aerial vehicle (UAV) transition section integrated anti-interference smooth switching control method of the present invention includes the following steps:

[0007] S1. Establish a longitudinal nonlinear dynamic model of the transition section of the tilt-rotor UAV, and select different control points in the tilt corridor for balancing to obtain a switching linear system model.

[0008] S2, introducing an external system to represent the disturbance input generated by the rotor wake;

[0009] S3, Design a disturbance observer to estimate disturbances from external systems;

[0010] S4. Design a reference model to transform the longitudinal control problem of the transition section into a comprehensive anti-disturbance control problem of the switching linear system, and further design a comprehensive anti-disturbance control law.

[0011] S5 employs a smooth interpolation control strategy to suppress state chattering caused by controller switching.

[0012] Furthermore, in step S1, the longitudinal nonlinear dynamic model of the transition section of the tilt-rotor UAV is expressed as follows:

[0013]

[0014] Where, [u,w,q,θ] T These are the four states of the system, where T represents matrix transpose; u is the forward velocity along the x-axis of the aircraft, w is the longitudinal velocity along the y-axis of the aircraft, q is the pitch velocity, and θ is the pitch angle; δ f For longitudinal periodic pitch, δ g For the total distance, δ e β is the rudder deflection angle. M Let f be the tilt angle; f() is a nonlinear function that satisfies kinematic and dynamic relationships.

[0015] Define a switching signal σ(t) that satisfies the modal-dependent residence time constraint: [0,∞)→N={1,2,…,n}, where n is the number of subsystems in the switching linear system model; for σ(t)=i, i∈N and any t y >t x ≥0, satisfying:

[0016] N i (t x ,t y )≤1+T i (t x ,t y ) / τ di

[0017] Where, Ni (t x ,t y T represents the total number of times the system is switched to the i-th subsystem. i (t x ,t y ) indicates that the i-th subsystem is in [t x ,t y The total running time within the interval is τ. di >0 is called the dwell time of modality dependence;

[0018] Furthermore, by selecting n sets of control points in the tilt corridor for balancing, the expression for the i-th subsystem of the switching linear system model is obtained as follows:

[0019]

[0020] Among them, x(t)=[Δu,Δw,Δq,Δθ] T These are the four state variables of the system: Δu is the increment of the forward velocity along the x-axis of the aircraft deviating from the state at the selected trim point; Δw is the increment of the longitudinal velocity along the y-axis of the aircraft deviating from the state at the selected trim point; Δq is the increment of the pitch angular velocity deviating from the state at the selected trim point; and Δθ is the increment of the pitch angle deviating from the state at the selected trim point. c (t)=[Δδ f ,Δδ g ,Δδ e ] T Δδ is the control input of the system. f Δδ represents the increment of the state at the selected balancing point due to the longitudinal periodic pitch deviation. g Δδ represents the increment of the total distance from the state at the selected balancing point. e ρ(t) represents the increment of the rudder deflection angle from the state at the selected trim point; ρ(t) represents the disturbance input from the external system generated by the rotor wake; ω1(t) is the disturbance input from the tiltrotor UAV; A i B i H i These are the constant matrices obtained by linearizing around the selected balancing point.

[0021] Furthermore, in step S2, an external system is introduced to represent the disturbance input generated by the rotor wake. For i∈N, the specific form of the external system is:

[0022]

[0023] ρ(t)=M i φ(t)

[0024] Where φ(t) represents the state of the external system; ω2(t) is the disturbance input acting on the external system, excluding the disturbance input generated by the rotor wake; C i G i M i These are constant matrices with appropriate dimensions.

[0025] Furthermore, in step S3, the expression for the disturbance observer is as follows:

[0026]

[0027]

[0028]

[0029] Where ψ(t) represents the system state of the perturbation observer; and Let represent the estimates of φ(t) and ρ(t) respectively; L is the perturbation observer gain;

[0030] The disturbance observation error is defined as:

[0031]

[0032] Furthermore, the differential of the perturbation observation error is obtained. The expression:

[0033]

[0034] Furthermore, in step S4, the expression for the reference model is as follows:

[0035]

[0036] Where, x r (t)=[Δu r ,Δw r ,Δq r ,Δθ r ] T For the ideal state that needs to be tracked, Δu r Δw is the increment of the ideal forward velocity along the x-axis of the aircraft from the state at the selected trim point. r Δq represents the increment of the ideal longitudinal velocity along the y-axis of the aircraft's coordinate system from the state at the selected balancing point. r Δθ is the increment of the ideal pitch angular velocity deviating from the state at the selected trim point. r The increment of the ideal pitch angle deviates from the state at the selected trim point; r(t) is a bounded reference input with appropriate dimension used to generate the ideal state trajectory; A r Br These are constant matrices with appropriate dimensions;

[0037] For the i-th subsystem of the switching linear system model, the control law is designed as follows:

[0038]

[0039] Among them, e a (t)=x(t)-x r (t) represents the state tracking error; K ei (t), K ci K ri These are the gain matrices of the controller to be designed;

[0040] Furthermore, the state tracking error e is obtained. a The expression for the differential of (t):

[0041]

[0042] The controller expression u c (t) Substitute From this, we obtain the expression for the tracking error system:

[0043]

[0044] Select K ci K ri Matrix, satisfying A i +B i K ci -A r =0 and B i K ri -B r If = 0 is true, then the tracking error system simplifies to:

[0045]

[0046] Furthermore, the error system state e(t) is defined as [e a (t),e b (t)] T And the perturbation input ω(t)=[ω1(t),ω2(t)] T Thus, the expression for the error system is obtained:

[0047]

[0048] in, and D i These are the coefficient matrices, each with its own expression:

[0049]

[0050] Furthermore, in step S5, t is used s Indicates the time when the s-th switch occurs, t s+1 Indicates the time when the (s+1)th handover occurs; introduces the transition interval [t]. s ,t s +τ h ), τ h >0 is a constant value for σ(t) s )=i∈N,τ h Less than the corresponding residence time τ di Furthermore, the interval [t] s ,t s+1 ) divided into [t s ,t s,0 )∪[t s,0 ,t s+1 ), where t s,0 =t s +τ h ;

[0051] Furthermore, a smooth interpolation strategy is designed to adjust the controller gain K. ei (t) is processed so that K ei (t)=U i (t)T i -1 (t); Assume there exists a coefficient matrix T. i >0 and U i Where i∈N; then for (i,j)∈N×N, i≠j, and σ(0)=m∈N, matrix U i (t) and T i The expression for (t) is:

[0052]

[0053]

[0054] Where, α(t)=(tt) s ) / τ h ; Coefficient matrices T i and U i Solve using linear matrix inequalities;

[0055] Selecting multiple Lyapunov functions:

[0056] V i (e(t))=e T (t)P i (t)e(t)

[0057] in, P represents the Lyapunov matrix of the i-th subsystem. 2i It is a positive definite matrix, and P 1i (t)=T i -1 (t);

[0058] To ensure the asymptotic stability of the error system and to have H ∞ Overall anti-interference performance, for the interval [t] s ,t s,0 The following inequalities must be satisfied:

[0059]

[0060] For the interval [t] s,0 ,t s+1 If the following inequality is satisfied, then the following inequality must be met:

[0061]

[0062] Where, η ui and η si γ and γ represent the decay rate of the multi-Lyapunov function in the corresponding time interval; γ>0 represents the L2 gain level of the system.

[0063] Furthermore, for σ(t) s )=i, i∈N,σ(t s - )=j, j∈N and i≠j; through K ei (t)=U i (t)T i -1 The processing of (t) requires satisfying the following linear matrix inequality:

[0064]

[0065]

[0066]

[0067] in,

[0068]

[0069]

[0070] Φ Aij =He{A i T i +B i U i}+η si T i

[0071] Furthermore, for a square matrix X, He{X} = X + X T ; I represents the identity matrix;

[0072] So, when the switching signal σ(t) s Satisfying the modal-dependent residence time τ di >τ h If the constraint is met, the error system is said to be asymptotically stable and has an L2 gain not exceeding the following equation.

[0073]

[0074] in, η u =max i∈N (η ui ), ν=exp{∑ i∈N (η si -η ui )τ h};

[0075] By solving the above linear matrix inequalities, we can further obtain matrix U. i (t) and T i (t); according to K ei (t)=U i (t)T i -1 (t), A i +B i K ci =A r And B i K ri =B r Solving for the controller gain matrix K ei (t), K ci and K ri .

[0076] Compared with the prior art, the significant advantages of this invention are as follows:

[0077] 1. This invention describes the disturbance input generated by the rotor wake by introducing an external system, and simultaneously designs a disturbance observer and H... ∞ The comprehensive anti-interference control scheme can suppress the effects of various types of disturbances;

[0078] 2. By designing a smooth interpolation control strategy, this invention can make the control input smoother, effectively suppress system state jumps, and make the transition process of tiltrotor UAVs more stable and safe. Attached Figure Description

[0079] Figure 1A schematic diagram of the tilt corridor for a tilt-rotor UAV;

[0080] Figure 2 This is a structural diagram of the present invention;

[0081] Figure 3(a) is a graph showing the forward flight speed response curve of the tiltrotor UAV during the transition phase obtained by using the integrated anti-interference control method of the present invention.

[0082] Figure 3(b) is a magnified view of part A in Figure 3(a);

[0083] Figure 4(a) is a graph showing the longitudinal velocity response curve of the transition section of the tiltrotor UAV obtained by using the integrated anti-interference control method of the present invention.

[0084] Figure 4(b) is a magnified view of part B in Figure 4(a);

[0085] Figure 5(a) shows the pitch angular velocity response curve of the transition section of the tilt rotor UAV obtained by using the integrated anti-interference control method of the present invention.

[0086] Figure 5(b) is a magnified view of C in Figure 5(a);

[0087] Figure 6 The diagram shows the pitch angle response curve of the transition section of a tilt-rotor UAV obtained using the integrated anti-interference control method of this invention. Detailed Implementation

[0088] 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.

[0089] During the transition phase of a tiltrotor aircraft, the rotation of the nacelle causes continuous changes in the aerodynamic shape. The strong nonlinear characteristics and drastic variations in aerodynamic parameters make it difficult for a single model to accurately describe the system's flight dynamics. However, using switched linear system modeling effectively solves this problem by leveraging the multi-subsystem description of the switched system. Therefore, compared to conventional nonlinear modeling, this invention, based on switched linear system modeling, can more accurately describe the flight dynamics of the tiltrotor UAV during the transition phase and simplifies controller design.

[0090] The integrated anti-interference control method for the transition section of a tilt-rotor unmanned aerial vehicle (UAV) of the present invention specifically includes the following steps:

[0091] Step 1: First, establish a longitudinal nonlinear dynamic model of the transition section of the tiltrotor UAV.

[0092] The modeling process employs a split-body modeling method, separately modeling the rotor, wing, fuselage, and vertical tail. Furthermore, assuming no lateral motion, it is simplified to a longitudinal dynamic model. Based on the Euler equations of dynamics, the longitudinal nonlinear dynamic model can be expressed as:

[0093]

[0094]

[0095]

[0096]

[0097] Where g is the acceleration due to gravity, m g For the mass of the aircraft; F Xr F Xw F Xf F Xt These represent the aerodynamic forces generated by the rotor, wing, fuselage, and vertical tail along the x-axis, respectively; F zr F zw F zf F zt These represent the aerodynamic forces generated by the rotor, wing, fuselage, and vertical tail in the z-axis direction, respectively; I y M represents the moment of inertia about the y-axis. r M w M f M t These represent the aerodynamic moments generated by the rotor, wing, fuselage, and vertical tail in the z-axis direction, respectively.

[0098] The established longitudinal nonlinear dynamic model can be simplified as follows:

[0099]

[0100] Where, [u,w,q,θ] T These are the four states of the system, where T represents matrix transpose; u is the forward velocity along the x-axis of the aircraft, w is the longitudinal velocity along the y-axis of the aircraft, q is the pitch velocity, and θ is the pitch angle; δ f For longitudinal periodic pitch, δ g For the total distance, δ e β is the rudder deflection angle. M Let f be the tilt angle; f() is a nonlinear function that satisfies kinematic and dynamic relationships.

[0101] Define a switching signal σ(t) satisfying the modal-dependent residence time constraint: [0,∞)→N={1,2,…,n}, where n is the number of subsystems in the switching linear system model. For σ(t)=i, i∈N and any ty >t x ≥0, satisfying:

[0102] N i (t x ,t y )≤1+T i (t x ,t y ) / τ di (6)

[0103] Where, N i (t x ,t y T represents the total number of times the system is switched to the i-th subsystem. i (t x ,t y ) indicates that the i-th subsystem is in [t x ,t y The total running time within the interval is τ. di >0 is called the dwell time of modality dependence;

[0104] Furthermore, by selecting n sets of control points in the tilt corridor for balancing, the expression for the i-th subsystem of the switching linear system model is obtained as follows:

[0105]

[0106] Among them, x(t)=[Δu,Δw,Δq,Δθ] T These are the four state variables of the system: Δu is the increment of the forward velocity along the x-axis of the aircraft deviating from the state at the selected trim point; Δw is the increment of the longitudinal velocity along the y-axis of the aircraft deviating from the state at the selected trim point; Δq is the increment of the pitch angular velocity deviating from the state at the selected trim point; and Δθ is the increment of the pitch angle deviating from the state at the selected trim point. c (t)=[Δδ f ,Δδ g ,Δδ e ] T Δδ is the control input of the system. f Δδ represents the increment of the state at the selected balancing point due to the longitudinal periodic pitch deviation. g Δδ represents the increment of the total distance from the state at the selected balancing point. e ρ(t) represents the increment of the rudder deflection angle from the state at the selected trim point; ρ(t) represents the disturbance input from the external system generated by the rotor wake; ω1(t) is the disturbance input acting on the tiltrotor UAV, which may be wind disturbance, etc.; A i B i H i These are the constant matrices obtained by linearizing around the selected balancing point.

[0107] Step 2: Introduce an external system to represent the disturbance input generated by the rotor wake. For i∈N, the specific form of the external system is:

[0108]

[0109] ρ(t)=M i φ(t) (9)

[0110] Where φ(t) represents the state of the external system; ω2(t) is the disturbance input acting on the external system (excluding the disturbance input generated by the rotor wake); ρ(t) represents the disturbance input generated by the rotor wake; C i G i M i These are constant matrices with appropriate dimensions.

[0111] Step 3, the expression for the disturbance observer is as follows:

[0112]

[0113]

[0114]

[0115] Where ψ(t) represents the system state of the perturbation observer; and Let represent the estimates of φ(t) and ρ(t) respectively; L is the perturbation observer gain;

[0116] The disturbance observation error is defined as:

[0117]

[0118] Furthermore, the differential of the perturbation observation error is obtained. The expression:

[0119]

[0120] Step 4, the designed reference model, is expressed as follows:

[0121]

[0122] Where, x r (t)=[Δu r ,Δw r ,Δq r ,Δθ r ] T For the ideal state that needs to be tracked, Δu rΔw is the increment of the ideal forward velocity along the x-axis of the aircraft from the state at the selected trim point. r Δq represents the increment of the ideal longitudinal velocity along the y-axis of the aircraft's coordinate system from the state at the selected balancing point. r Δθ is the increment of the ideal pitch angular velocity deviating from the state at the selected trim point. r The increment of the ideal pitch angle deviates from the state at the selected trim point; r(t) is a bounded reference input with appropriate dimension used to generate the ideal state trajectory; A r B r These are constant matrices with appropriate dimensions;

[0123] For the i-th subsystem of the switching linear system model, the control law is designed as follows:

[0124]

[0125] Among them, e a (t)=x(t)-x r (t) represents the state tracking error; K ei (t), K ci K ri These are the gain matrices of the controller to be designed;

[0126] Furthermore, the state tracking error e is obtained. a The expression for the differential of (t):

[0127]

[0128] The controller expression u c (t) Substitute From this, we obtain the expression for the tracking error system:

[0129]

[0130] Select K ci K ri Matrix, satisfying A i +B i K ci -A r =0 and B i K ri -B r If = 0 is true, then the tracking error system simplifies to:

[0131]

[0132] Furthermore, the error system state e(t) is defined as [e a (t),e b (t)] TAnd the perturbation input ω(t)=[ω1(t),ω2(t)] T Thus, the expression for the error system is obtained:

[0133]

[0134] in, and D i These are the coefficient matrices, each with its own expression:

[0135]

[0136] Step 5, use t s Indicates the time when the s-th switch occurs, t s+1 Indicates the time when the (s+1)th handover occurs; introduces the transition interval [t]. s ,t s +τ h ), τ h >0 is a constant value for σ(t) s )=i∈N,τ h Less than the corresponding residence time τ di Furthermore, the interval [t] s ,t s+1 ) divided into [t s ,t s,0 )∪[t s,0 ,t s+1 ), where t s,0 =t s +τ h ;

[0137] Furthermore, a smooth interpolation strategy is designed to adjust the controller gain k. ei (t) is processed so that K ei (t)=U i (t)T i -1 (t); Assume there exists a coefficient matrix T. i >0 and U i Where i∈N; then for (i,j)∈N×N, i≠j, and σ(0)=m∈N, matrix U i (t) and T i The expression for (t) is:

[0138]

[0139]

[0140] Where, α(t)=(tt) s ) / τ h ; Coefficient matrices T i and Ui It can be solved using subsequent linear matrix inequalities;

[0141] Selecting multiple Lyapunov functions:

[0142] V i (e(t))=e T (t)P i (t)e(t) (23)

[0143] in, P represents the Lyapunov matrix of the i-th subsystem. 2i It is a positive definite matrix, and P 1i (t)=T i -1 (t); * denotes a matrix that can be obtained by symmetry.

[0144] To ensure the asymptotic stability of the error system and to have H ∞ Overall anti-interference performance, for the interval [t] s ,t s,0 The following inequalities must be satisfied:

[0145]

[0146] For the interval [t] s,0 ,t s+1 If the following inequality is satisfied, then the following inequality must be met:

[0147]

[0148] Where, η ui and η si γ and γ represent the decay rate of the multi-Lyapunov function in the corresponding time interval; γ>0 represents the L2 gain level of the system.

[0149] Furthermore, for σ(t) s )=i, i∈N,σ(t s - )=j, j∈N and i≠j; through K ei (t)=U i (t)T i -1 The processing of (t) requires satisfying the following linear matrix inequality:

[0150]

[0151]

[0152]

[0153] in,

[0154]

[0155]

[0156] Φ Aij =He{A i T i +B i U i}+η si T i

[0157] Furthermore, for a square matrix X, He{X} = X + X T ; I represents the identity matrix;

[0158] So, when the switching signal σ(t) s Satisfying the modal-dependent residence time τ di >τ h If the constraint is met, the error system is said to be asymptotically stable and has an L2 gain not exceeding the following equation.

[0159]

[0160] in, η u =max i∈N (η ui ), v = exp{∑ i∈N (η si -η ui )τ h};

[0161] By solving the above linear matrix inequalities, we can further obtain matrix U. i (t) and T i (t); according to K ei (t)=U i (t)T i -1 (t), A i +B i K ci =A r And B i K ri =B r Solving for the controller gain matrix K ei (t), K ci and K ri .

[0162] To verify the effectiveness of the present invention in the comprehensive anti-interference control of the tilt rotor transition section, the following simulation experiment was conducted.

[0163] exist Figure 1 In the tilting corridor shown, the nacelle angle is the complementary angle of the tilting angle; control points near tilting angles of 0°, 12°, 35°, 52°, and 90° are selected to balance the system; and a structural diagram of the control method used in this example is constructed, as shown below. Figure 2 As shown.

[0164] Assuming the entire tilting process ends in 15 seconds, the switching signal σ(t) is designed as follows: when t∈[0,2), σ(t)=1; when t∈[2,6), σ(t)=2; when t∈[6,9), σ(t)=3; when t∈[9,12), σ(t)=4; when t∈[12,15), σ(t)=5.

[0165] Figures 3(a)-6 In the comparison method, the traditional switching control method is represented, which removes the smooth interpolation control strategy designed in this invention. Different controller gain matrices are obtained by solving the corresponding matrix linear inequalities. The rest, including the reference system and coefficient matrix, remains consistent with the method proposed in this invention. Simulations were performed to obtain comparison diagrams of the state response curves of this invention and the comparison method, as shown in Figures 3(a), 4(a), and 5(a). Figure 6 As shown.

[0166] It can be seen that, Figure 3(a) , 3(b) Both methods showed good control over the forward velocity u during the transition phase of the tiltrotor UAV, but the method proposed in this invention can better track the ideal trajectory. Figures 4(a) and 4(b) show the state response curves for the longitudinal velocity w, and Figures 5(a) and 5(b) show the state response curves for the pitch angular velocity q. Figure 6 The figure shows the state response curve for the pitch angle θ. As can be seen from the curves above, the method proposed in this invention enables the tiltrotor UAV to achieve a smooth climb in longitudinal velocity and a stable change in pitch angle throughout the entire transition phase. Furthermore, it allows the system state to track the ideal trajectory, thus achieving a smooth transition in the tiltrotor UAV's flight mode. Simultaneously, by comparing the two methods in the simulation figures, it can be seen that the smooth interpolation H designed by the method proposed in this invention... ∞ The control strategy can effectively suppress state transitions during switching, improve system transient performance, and achieve better control results. The simulation results above fully demonstrate that the integrated anti-interference smooth control method for the transition phase of a tiltrotor UAV can effectively solve the problem of flight mode switching in tiltrotor aircraft.

Claims

1. A comprehensive anti-interference smooth switching control method for the transition section of a tiltrotor unmanned aerial vehicle, characterized in that, Includes the following steps: S1. Establish a longitudinal nonlinear dynamic model of the transition section of the tilt-rotor UAV, and balance it by selecting different control points in the tilt corridor to obtain a switching linear system model; define a switching signal that satisfies the modal-dependent dwell time constraint. , To switch the number of subsystems in a linear system model; for , And any ,satisfy: ， in, Switching to the first The total number of subsystems Indicates the first Subsystems in The total running time within the interval, then The residence time is called modality-dependent; S2, introducing an external system to represent the disturbance input generated by the rotor wake; S3, Design a disturbance observer to estimate disturbances from external systems; S4. Design a reference model to transform the longitudinal control problem of the transition section into a comprehensive anti-disturbance control problem of the switching linear system, and further design a comprehensive anti-disturbance control law. S5, a smooth interpolation control strategy is designed to suppress state chattering caused by controller switching; In step S3, the expression for the disturbance observer is as follows: ， ， ， in, Indicates the system state of the disturbance observer; and They represent respectively to and The estimate, Indicates the state of the external system. This represents the disturbance input generated by the rotor wake in the external system; For the perturbation observer gain; These are the four state variables of the system. For along the body The increment of the forward velocity along the coordinate axis deviating from the state at the selected trim point. For the body coordinates The increment of the longitudinal velocity in the axial direction deviating from the state at the selected balancing point. This is the increment of the pitch angular velocity deviation from the state at the selected trim point. The increment of the pitch angle deviation from the state at the selected trim point; For the system's control input, This represents the increment of the state at the selected balancing point due to the longitudinal periodic pitch deviation. This is the increment of the total pitch deviation from the state at the selected balancing point. This is the increment of the rudder deflection angle from the state at the selected trim point; , These are the constant matrices obtained by linearizing around the selected balancing point; A constant matrix of appropriate dimension; T denotes matrix transpose; It is a constant matrix with appropriate dimensions; Define the perturbation observation error for: , Furthermore, the differential of the perturbation observation error is obtained. The expression: ， in, It is the disturbance input from the tiltrotor drone. It is a disturbance input that acts on the external system; This is the constant matrix obtained by linearizing the matrix near the selected balancing point. A constant matrix with appropriate dimensions; In step S4, the expression of the reference model is as follows: ， in, For the ideal state that needs to be tracked, For along the body The increment of the ideal forward velocity in the coordinate axis direction from the state at the selected trim point. For the body coordinates The increment of the ideal longitudinal velocity in the axial direction from the state at the selected trim point. The increment of the ideal pitch angular velocity deviating from the state at the selected trim point. The increment of the deviation of the ideal pitch angle from the state at the selected trim point; A bounded reference input with appropriate dimensions is used to generate an ideal state trajectory; , These are constant matrices with appropriate dimensions; For the switching linear system model, the first The control law for each subsystem is designed as follows: ， in, This refers to the state tracking error; , , These are the gain matrices of the controller to be designed; Furthermore, the state tracking error is obtained. The expression for the differential: ， controller expression Substitution From this, we obtain the expression for the tracking error system: ， Select , Matrix, satisfying as well as If true, then the tracking error system simplifies to: ； Furthermore, define the error system state. and disturbance input Thus, the expression for the error system is obtained: ， in, and These are the coefficient matrices, each with its own expression: , ； In step S5, using Indicates the occurrence of the first The moment of switching Indicates the first The timing of the handover; introduction of a transition interval. , It is a constant value, for , Less than the corresponding stay time Furthermore, the interval Divided into ,in, ; Furthermore, a smooth interpolation strategy is designed to affect the controller gain. Processing is performed to make Assume there exists a coefficient matrix. and ,in So for , ,and ,matrix and The expression is: , , in, Coefficient matrices and Solve using linear matrix inequalities; Selecting multiple Lyapunov functions: ， in, Representing the Lyapunov matrices of each subsystem It is a positive definite matrix, and ; To ensure the asymptotic stability of the error system and have Overall anti-interference performance, for interval The following inequalities must be satisfied: ， For interval Then the following inequality must be satisfied: ， in, and These represent the decay rates of the Lyapunov function within the corresponding time intervals; Representative system Gain level; Furthermore, for , , , and ;pass The processing of this requires satisfying the following linear matrix inequality: ， ， ， in, ， ， ， Furthermore, regarding the square array , ; Represents the identity matrix; So, when switching signals Satisfying modal dependency residence time If the constraint is met, then the error system is said to be asymptotically stable and has a constraint not exceeding the following: Gain : ， in, , , ; By solving the above linear matrix inequalities, we can further obtain the matrix. and ;according to , as well as Solve for the controller gain matrix. , and .

2. The comprehensive anti-interference smooth switching control method for the transition section of a tiltrotor UAV according to claim 1, characterized in that, In step S1, the longitudinal nonlinear dynamic model of the transition section of the tilt-rotor UAV is expressed as follows: ， in, These are the four states of the system. Indicates matrix transpose; For along the body Forward velocity along the coordinate axis direction, For the body coordinates Longitudinal velocity in the axial direction, The pitch angular velocity, The pitch angle; For longitudinal periodic pitch, For the total distance, This refers to the rudder deflection angle; The tilt angle; To satisfy the nonlinear function of kinematic and dynamic relationships; Furthermore, in the tilting corridor, select The group of control points is balanced to obtain the first step of the switching linear system model. The expression for each subsystem is: ， in, This represents the disturbance input generated by the rotor wake in the external system.

3. The comprehensive anti-interference smooth switching control method for the transition section of a tiltrotor UAV according to claim 2, characterized in that, In step S2, an external system is introduced to represent the disturbance input generated by the rotor wake. The specific form of the external system is as follows: ， ， in, Indicates the state of the external system; It refers to the disturbance input acting on the external system, excluding the disturbance input generated by the rotor wake.

Images