A robust H∞ control method and system for target drone flight at high angle of attack

Abstract

This application provides a robust H∞ control method and system for target drone flight at high angle of attack. The method includes: Step 1, obtaining the target drone speed, target drone angular velocity, target drone roll angle, and target drone airflow angle; Step 2, setting the constraint domain of the closed-loop poles of the robust control system for target drone flight at high angle of attack; Step 3, setting the tracking error model and target drone uncertainty structure of the robust control system for target drone flight at high angle of attack; Step 4, setting the robust H∞ control method for target drone flight at high angle of attack. ∞ The control method uses a system of linear matrix inequalities for the control parameters; Step 5: Set the objective of solving the system of linear matrix inequalities for the control parameters, and use the linear matrix inequality toolbox to solve the control parameters; Apply the parameters to the target drone's robust control system for high angle-of-attack flight to obtain the target drone's robust H for high angle-of-attack flight. ∞ The controller output; Step 6: Apply the values ​​of the aileron, elevator and rudder output by the controller to the target drone to realize the target drone tracking the high angle of attack command, sideslip angle command and track roll angle command.

Description

Technical Field

[0001] This invention belongs to the field of target drone flight control, and specifically relates to a robust H∞ control method and system for target drone flight at high angle of attack. Background Technology

[0002] To meet the requirements of target drone testing with a wide speed range, large flight envelope, and high maneuverability, target drones are required to possess the ability to decelerate over a wide range, turn in a short time, and turn with a small radius. This enables them to quickly change their nose direction and evade missile attacks, thus significantly testing the combat effectiveness of new types of weapons and equipment. Maneuvering is one of the ways for target drones to achieve these capabilities.

[0003] High angle-of-attack flight of target drones is fundamental to maneuvering flight; therefore, studying high angle-of-attack flight of target drones is essential. However, when a target drone is at a high angle of attack, its aerodynamic characteristics undergo drastic changes, mainly in the following aspects:

[0004] Nonlinear and unsteady aerodynamics: When a target drone is at a high angle of attack, its wings and fuselage will experience airflow separation and vortex breaking as the angle of attack and Mach number increase, leading to abrupt changes in its aerodynamic characteristics, exhibiting unsteady and strongly nonlinear behavior. Simultaneously, its dynamics and kinematic models themselves possess strong nonlinear characteristics. The nonlinear and unsteady aerodynamics of the target drone are extremely detrimental to its flight control at high angles of attack. The abrupt changes in aerodynamic characteristics mean that the target drone no longer possesses a typical trim state, and the assumptions of the traditional small-disturbance linearization model no longer hold. Since the aerodynamic model is the foundation for designing and analyzing controllers, linear controllers developed based on linear models will also fail. Therefore, designing a fast, accurate, and stable advanced nonlinear controller to achieve high angle-of-attack flight for target drones is currently an important research direction.

[0005] Uncertainties and disturbances: At high angles of attack, the control surfaces of the target drone exhibit reduced effectiveness due to airflow separation, leading to uncertainties in the input matrix. Complex external flight environments, such as gusts and turbulence, as well as input gain matrices of unknown magnitude, also introduce uncertain disturbances to the target drone. Furthermore, the current modeling theory for high angle-of-attack target drones is incomplete, resulting in significant deviations in the dynamic models themselves, including inaccurate aerodynamic data and the presence of unmodeled dynamics. Therefore, improving the robustness and anti-interference capability of the nonlinear controller for the target drone model is crucial for achieving high angle-of-attack flight of the target drone.

[0006] Coupling Effect: For conventional aircraft, the angle-of-attack range within their flight envelope is relatively small, and the aerodynamic data can be considered quasi-constant. Therefore, without considering the effects of external interference and brake delay characteristics, a dynamic inverse controller can accurately eliminate the coupling between the target drone's roll and yaw directions. However, for target drones flying at high angles of attack, their aerodynamic data changes with the angle of attack, especially undergoing abrupt changes at high angles of attack. Therefore, under the influence of nonlinear and unsteady aerodynamic data, coupled with inherent modeling errors, the coupling between the target drone's roll and yaw directions can no longer be accurately eliminated, leading to uncertainties in roll and yaw directions. Therefore, how to suppress or compensate for the effects of the coupling between the target drone's roll and yaw directions and design an accurate and stable controller is an inevitable challenge in achieving high angle-of-attack flight for target drones.

[0007] Due to the aforementioned characteristics, the lift coefficient of a target drone decreases as the angle of attack increases at high angles of attack, which is the opposite of the trend at low angles of attack. If the target drone cannot recover accurately at the appropriate time, it will enter a deep stall state, which can easily lead to serious consequences. If the target drone sideslips in a stall state, it will generate spontaneous yaw motion, and oscillate in the pitch, roll, and yaw directions, entering a spin state, which seriously affects the stability and safety of the target drone.

[0008] To address the aforementioned issues, this invention provides a robust, highly anti-interference, and high-precision nonlinear robust control system that enables the target drone to fly well under high angle-of-attack conditions. Summary of the Invention

[0009] The purpose of this invention is to provide a target drone with robust H-axis flight at high angles of attack. ∞ The control method can enable the target drone to quickly and accurately track large angle of attack, sideslip angle and track roll angle commands even when the target drone has model uncertainty, severe unsteadiness and nonlinearity, as well as severe external gust interference and yaw and roll direction coupling. The closed-loop poles of the target drone are within the preset limit domain.

[0010] The technical solution of this invention:

[0011] Firstly, this application provides a target drone with robust H-axis flight at high angle of attack. ∞ Control method, the method being applied to a robust control system for high angle-of-attack flight of a target drone, the method comprising:

[0012] Step 1: Obtain the target drone's speed, angular velocity, roll angle, and airflow angle using an accelerometer, gyroscope, GPS, and airflow angle sensor; the airflow angle consists of the target drone's angle of attack and sideslip angle.

[0013] Step 2: Set the limiting domain of the closed-loop poles of the target drone's high angle-of-attack flight robust control system;

[0014] Step 3: Set up the tracking error model and target uncertainty structure for the target drone's robust control system for high angle of attack flight;

[0015] Step 4: Set the target drone's high angle-of-attack flight robustness H ∞ The control method comprises a set of linear matrix inequalities for control parameters; the set of linear matrix inequalities for control parameters consists of parameter-solved linear matrix inequalities and parameter-restricted linear matrix inequalities.

[0016] Step 5: Set the objective of solving the system of linear matrix inequalities for control parameters. Use the Linear Matrix Inequality Toolbox (LMI) in Matlab simulation software to calculate the objective and obtain the target drone's robustness H at high angle of attack. ∞ Parameters of the control method; robustness of the target drone at high angle of attack. ∞ The parameters of the control method are applied to the target drone's robust control system for high angle-of-attack flight to obtain the target drone's robustness H for high angle-of-attack flight. ∞ The controller output; the target drone's robustness to high angle-of-attack flight H ∞ The controller outputs the target drone's aileron deflection value, target drone elevator deflection value, and target drone rudder deflection value.

[0017] Step 6: Apply the target drone's aileron deflection value, elevator deflection value, and rudder deflection value to the target drone to enable the target drone to quickly, accurately, and simultaneously track high angle of attack commands, sideslip angle commands, and track roll angle commands; the high angle of attack command is a command with an angle of attack greater than 25°.

[0018] Furthermore, step 1 includes:

[0019] Step 11: Obtain the target velocity V in the target drone airflow coordinate system using an accelerometer; the target drone airflow coordinate system has three axes: x, y, and z, with the x, y, and z axes pointing forward, right, and downward, respectively; the direction of the target velocity V is the x-axis direction;

[0020] Step 12: Obtain the target drone's angle of attack α and sideslip angle β using the airflow angle sensor;

[0021] Step 13: Obtain the target angular velocity ω in the target body coordinate system using the gyroscope. b =[ω x ω y ω z ] T The target body coordinate system has three axes: x, y, and z. The x, y, and z axes point forward, right, and downward, respectively. The ω... x ω y and ω z These are the target drone angular velocities along the x-axis, y-axis, and z-axis in the target drone's body coordinate system, respectively.

[0022] Step 14: Obtain the target drone's roll angle μ from GPS.

[0023] Furthermore, step 2 includes:

[0024] Step 21: Set the closed-loop pole constraint region of the target drone's high angle-of-attack flight robust control system to a circular region located to the left of the imaginary axis of the Laplace-variant complex plane; the specific form of the closed-loop pole constraint region φ is as follows:

[0025]

[0026] Where σ>0, κ>0, -σ is the center position of the circular domain, κ is the radius of the circular domain, and s is the complex domain of the Laplace transform. It is the conjugate of s.

[0027] Furthermore, step 3 includes:

[0028] Step 31: Set the desired angle of attack α for the target drone d Expected target drone sideslip angle β d Expected target drone roll angle μ d ;

[0029] Step 32: Set the airflow state error e1 for the target drone; the airflow state error e1 for the target drone is the desired angle of attack α of the target drone. d Expected target drone sideslip angle β d Expected target drone roll angle μ d The difference between the target drone's angle of attack α, sideslip angle β, and roll angle μ, and the specific form of the target drone's airflow state error e1, is: e1=[α d -αβ d -βμ d -μ] T ;

[0030] Step 33: Set the desired target angular velocity [ω] xd ω yd ω zd ] T ;

[0031] Step 34: Set the target drone angular velocity error e2; the target drone angular velocity error e2 is the desired target drone angular velocity [ω]. xd ω yd ω zd ] T With the target angular velocity [ω x ω y ω z ] T The difference between them, the specific form of the target angular velocity error e2 is: e2=[ω xd -ω x ω yd-ω y ω zd -ω z ] T ;

[0032] Step 35: Set the tracking error model D of the target drone's high angle-of-attack flight robust control system; the specific form of the tracking error model D of the target drone's high angle-of-attack flight robust control system is as follows:

[0033]

[0034] Among them, the The total tracking error of the target drone. Let Φ(V) be the derivative of the total tracking error of the target drone with respect to time, Φ(V) be the target drone state matrix related to the target drone velocity V, Θ(V) be the target drone input matrix related to the target drone velocity V, ΔΦ and ΔΘ be the target drone uncertainties, d represent the external gust interference of the target drone, and u represent the target drone's robustness at high angle of attack. ∞ The controller output includes the target drone's aileron deflection λ. ail Target drone elevator deflection λ ele Target drone rudder offset λ rud ;

[0035] Step 36: Set the target uncertainty structure for the target uncertainty ΔΦ and ΔΘ as: [ΔΦΔΘ]=QΓT(t)[Γ -1 R1Γ -1 R2];

[0036] Wherein, matrices Q, R1, and R2 are known constant matrices of dimension 6×6, representing the structural information of uncertainties ΔΦ and ΔΘ; T(t) is a time-varying unknown matrix, representing the uncertainty of uncertainties ΔΦ and ΔΘ; Γ is a diagonal matrix, and I is the identity matrix; wherein the values ​​of matrices Q, R1, and R2 should and satisfy T(t)T T (t)≤I.

[0037] Furthermore, step 4 includes:

[0038] Step 41: Set the target drone's high angle-of-attack flight robustness H ∞ The linear matrix inequality for solving the parameters of the control method is:

[0039]

[0040] Wherein, the positive definite matrix Y, the diagonal matrix Γ, the matrix V, and the constant τ represent the target drone's robustness H at high angle of attack flight, which needs to be solved. ∞The parameters of the control method are: I is the identity matrix; diagonal matrices D and E are the error weight matrix and input weight matrix of the target drone's robust control system for high angle-of-attack flight, respectively; and ξ is the H value of the external gust disturbance d to the target drone's robust control system for high angle-of-attack flight. ∞ Norm index; the H ∞ The norm index ξ is in the range of 0 < ξ < 0.5;

[0041] Step 42: Set the target drone's high angle-of-attack flight robustness H ∞ The parameter constraint linear matrix inequality for the control method is:

[0042]

[0043] Wherein, W is the parameter constraint matrix, and I is the identity matrix;

[0044] Step 43: Solve the linear matrix inequalities with parameters and parameter constraints to obtain the target drone's high angle-of-attack flight robustness H. ∞ The control parameters of the control method are a set of linear matrix inequalities.

[0045] Furthermore, step 5 includes:

[0046] Step 51: Set the control parameters. The objective of solving the system of linear matrix inequalities is: minTrace(W)

[0047] Wherein, min represents the minimization operation, and Trace(W) represents the trace of the parameter constraint matrix W;

[0048] Step 52: Use the Linear Matrix Inequality Toolbox (LMI) in Matlab simulation software to calculate the target drone's high angle-of-attack flight robustness H with minTrace(W) as the target. ∞ The control parameter linear matrix inequalities of the control method are used to obtain the target drone's robustness H at high angle of attack. ∞ The parameters of the control method are the positive definite matrix Y, the diagonal matrix Γ, the matrix V, and the constant τ;

[0049] Step 53: Set the target drone's high angle-of-attack flight robustness H ∞ The controller output u is:

[0050] Obtain the target drone aileron deflection λ ail Target drone elevator deflection λ ele Target drone rudder offset λ rud .

[0051] Furthermore, the structure of the closed-loop pole constraint domain φ of the target drone's high angle-of-attack flight robust control system is as follows: Where s refers to the complex field of the Laplace transform. is the conjugate complex number of s, and the symmetric matrices Ω and Ξ are used to describe the shape of the closed-loop pole confinement domain;

[0052] The closed-loop pole-limiting domain of the target drone's high angle-of-attack flight robust control system can ensure that the system meets the requirements for steady-state and transient performance.

[0053] Set the closed-loop pole constraint region to a circular region, and set the matrix... matrix Wherein, σ>0, κ>0, -σ is the center position of the circular domain, and κ is the radius of the circular domain.

[0054] If the target drone's high angle-of-attack flight robust control system requires fast convergence speed and small steady-state error, then the range of parameter σ is: σ>10, and the range of parameter κ is: 0<κ<2. If the target drone's high angle-of-attack flight robust control system requires small overshoot and fast ascent speed, then the range of parameter σ is: 4<σ<8, and the range of parameter κ is: 4<κ<6.

[0055] Secondly, this application provides a robust control system for high angle-of-attack flight, the system including an accelerometer, a gyroscope, GPS, an airflow angle sensor, and a target drone high angle-of-attack flight robust H... ∞ Controller, ailerons, elevator, and rudder; where: the target drone's velocity V measured by the accelerometer, and the target drone's angular velocity ω measured by the gyroscope. b =[ω x ω y ω z ] T The target drone's roll angle μ measured by GPS, and the target drone's angle of attack α and sideslip angle β measured by the airflow angle sensor are the robustness H of the target drone's flight at high angles of attack. ∞ Controller inputs; target drone high angle-of-attack flight robustness H ∞ The controller outputs the deflection values ​​for the ailerons, elevator, and rudder.

[0056] Advantages of this invention:

[0057] This invention provides a target drone with robust H-axis flight at high angle of attack. ∞ The control method can achieve the following even when the target drone exhibits model uncertainty, severe unsteadiness and nonlinearity, as well as severe external gust interference and yaw / roll direction coupling:

[0058] To ensure the asymptotic stability of the target drone's robust control system during high angle-of-attack flight;

[0059] The target drone can quickly and accurately track commands for high angle of attack, sideslip angle, and track roll angle.

[0060] The control method is robust to the target drone model parameters and does not require precise model parameters of the target drone. It only needs to have an estimate of the range of parameter variation to realize the target drone tracking control command even when the model parameters are uncertain.

[0061] The control method is robust to external gust interference, achieving a transfer function H of the robust control system for high angle-of-attack flight of the target drone from external gust interference. ∞ The norm is less than the index value, meaning that external gust interference has little impact on the robust control system of the target drone at high angle of attack.

[0062] The target drone's high angle-of-attack flight robust control system is quadratic D-stable, meaning that the closed-loop poles of the target drone's high angle-of-attack flight robust control system are within a preset limit domain.

[0063] The target drone's robust control system for high angle-of-attack flight meets the requirements for both steady-state and transient performance. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of a robust control system for high angle-of-attack flight provided in this application;

[0065] Figure 2 This application provides a flowchart of the steps of a robust control system for high angle-of-attack flight.

[0066] Figure 3 This application provides a target drone with robust H-type flight at high angle of attack. ∞ A schematic diagram of the control method;

[0067] Figure 4 It is a target drone tracking angle of attack command response curve;

[0068] Figure 5 This is a graph showing the target drone's response to the sideslip angle command.

[0069] Figure 6 This is a curve showing the response of the target drone to the roll angle command. Specific Implementation

[0070] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0071] Example 1

[0072] This application provides a target drone with robust H-type high angle-of-attack flight. ∞ A control method, which is applied to a robust control system for high angle-of-attack flight.

[0073] like Figure 1 As shown, the target drone's robust control system for high angle-of-attack flight includes an accelerometer, gyroscope, GPS, airflow angle sensor, and target drone high angle-of-attack flight robust H... ∞Controller, ailerons, elevator, and rudder; where: the target drone's velocity V measured by the accelerometer, and the target drone's angular velocity ω measured by the gyroscope. b =[ω x ω y ω z ] T The target drone's roll angle μ measured by GPS, and the target drone's angle of attack α and sideslip angle β measured by the airflow angle sensor are the robustness H of the target drone's flight at high angles of attack. ∞ Controller inputs; target drone high angle-of-attack flight robustness H ∞ The controller outputs the deflection values ​​for the ailerons, elevator, and rudder.

[0074] like Figure 2 As shown, the target drone is robust to high angle-of-attack flight H ∞ Control methods include:

[0075] Step 1: Obtain the target drone's speed, angular velocity, roll angle, and airflow angle using an accelerometer, gyroscope, GPS, and airflow angle sensor; the airflow angle consists of the target drone's angle of attack and sideslip angle.

[0076] Specifically, step 1 includes:

[0077] Step 11: Obtain the target velocity V in the target drone airflow coordinate system using an accelerometer; the target drone airflow coordinate system has three axes: x, y, and z, with the x, y, and z axes pointing forward, right, and downward, respectively; the direction of the target velocity V is the x-axis direction;

[0078] Step 12: Obtain the target drone's angle of attack α and sideslip angle β using the airflow angle sensor;

[0079] Step 13: Obtain the target angular velocity ω in the target body coordinate system using the gyroscope. b =[ω x ω y ω z ] T The target body coordinate system has three axes: x, y, and z. The x, y, and z axes point forward, right, and downward, respectively. The ω... x ω y and ω z These are the target drone angular velocities along the x-axis, y-axis, and z-axis in the target drone's body coordinate system, respectively.

[0080] Step 14: Obtain the target drone's roll angle μ from GPS.

[0081] Step 2: Set the limiting domain of the closed-loop poles of the target drone's high angle-of-attack flight robust control system;

[0082] Specifically, step 2 includes:

[0083] Step 21: Set the closed-loop pole constraint region of the target drone's high angle-of-attack flight robust control system to a circular region located to the left of the imaginary axis of the Laplace-variant complex plane; the specific form of the closed-loop pole constraint region φ is as follows:

[0084]

[0085] Where σ>0, κ>0, -σ is the center position of the circular domain, κ is the radius of the circular domain, and s is the complex domain of the Laplace transform. It is the conjugate of the complex number s;

[0086] Step 22: If the target drone's high angle-of-attack flight robust control system requires fast convergence speed and small steady-state error, then the range of parameter σ is: σ>10, and the range of parameter κ is: 0<κ<2; If the target drone's high angle-of-attack flight robust control system requires small overshoot and fast ascent speed, then the range of parameter σ is: 4<σ<8, and the range of parameter κ is: 4<κ<6.

[0087] Step 23: The closed-loop pole restriction domain of the target drone's high angle-of-attack flight robust control system can ensure that the target drone's high angle-of-attack flight robust control system meets the requirements of steady-state performance and transient performance.

[0088] Step 3: Set up the tracking error model and target uncertainty structure for the target drone's robust control system for high angle of attack flight;

[0089] Specifically, step 3 includes:

[0090] Step 31: Set the desired angle of attack α for the target drone d Expected target drone sideslip angle β d Expected target drone roll angle μ d ;

[0091] Step 32: Set the airflow state error e1 for the target drone; the airflow state error e1 for the target drone is the desired angle of attack α of the target drone. d Expected target drone sideslip angle β d Expected target drone roll angle μ d The difference between the target drone's angle of attack α, sideslip angle β, and roll angle μ, and the specific form of the target drone's airflow state error e1, is: e1=[α d -αβ d -βμ d -μ] T ;

[0092] Step 33: Set the desired target angular velocity [ω] xd ω yd ω zd ] T ;

[0093] Step 34: Set the target drone angular velocity error e2; the target drone angular velocity error e2 is the desired target drone angular velocity [ω]. xd ω yd ω zd ] T With the target angular velocity [ω x ω y ω z ] T The difference between them, the specific form of the target angular velocity error e2 is: e2=[ω xd -ω x ω yd -ω y ω zd -ω z [ T ;

[0094] Step 35: Set the tracking error model D of the target drone's high angle-of-attack flight robust control system; the specific form of the tracking error model D of the target drone's high angle-of-attack flight robust control system is as follows:

[0095]

[0096] Among them, the The total tracking error of the target drone. Let Φ(V) be the derivative of the total tracking error of the target drone with respect to time, Φ(V) be the target drone state matrix related to the target drone velocity V, Θ(V) be the target drone input matrix related to the target drone velocity V, ΔΦ and ΔΘ be the target drone uncertainties, d represent the external gust interference of the target drone, and u represent the target drone's robustness at high angle of attack. ∞ The controller output includes the target drone's aileron deflection λ. ail Target drone elevator deflection λ ele Target drone rudder offset λ rud ;

[0097] Step 36: Set the target uncertainty structure for the target uncertainty ΔΦ and ΔΘ as: [ΔΦΔΘ]=QΓT(t)[Γ -1 R1Γ -1 R2];

[0098] Wherein, matrices Q, R1, and R2 are known constant matrices of dimension 6×6, representing the structural information of uncertainties ΔΦ and ΔΘ; T(t) is a time-varying unknown matrix, representing the uncertainty of uncertainties ΔΦ and ΔΘ; Γ is a diagonal matrix, and I is the identity matrix; wherein the values ​​of matrices Q, R1, and R2 should and satisfy T(t)T T (t)≤I.

[0099] Step 4: Set the target drone's high angle-of-attack flight robustness H ∞The control method comprises a set of linear matrix inequalities for control parameters; the set of linear matrix inequalities for control parameters consists of parameter-solved linear matrix inequalities and parameter-restricted linear matrix inequalities.

[0100] Specifically, step 4 includes:

[0101] Step 41: Set the target drone's high angle-of-attack flight robustness H ∞ The linear matrix inequality for solving the parameters of the control method is:

[0102]

[0103] Wherein, the positive definite matrix Y, the diagonal matrix Γ, the matrix V, and the constant τ represent the target drone's robustness H at high angle of attack flight, which needs to be solved. ∞ The parameters of the control method are: I is the identity matrix; diagonal matrices D and E are the error weight matrix and input weight matrix of the target drone's robust control system for high angle-of-attack flight, respectively; and ξ is the H value of the external gust disturbance d to the target drone's robust control system for high angle-of-attack flight. ∞ Norm index; the H ∞ The norm index ξ is in the range of 0 < ξ < 0.5;

[0104] Step 42: Set the target drone's high angle-of-attack flight robustness H ∞ The parameter constraint linear matrix inequality for the control method is:

[0105]

[0106] Wherein, W is the parameter constraint matrix, and I is the identity matrix;

[0107] Step 43: Solve the linear matrix inequalities with parameters and parameter constraints to obtain the target drone's high angle-of-attack flight robustness H. ∞ The control parameters of the control method are a set of linear matrix inequalities.

[0108] Step 5: Set the objective of solving the system of linear matrix inequalities for control parameters. Use the Linear Matrix Inequality Toolbox (LMI) in Matlab simulation software to calculate the objective and obtain the target drone's robustness H at high angle of attack. ∞ Parameters of the control method; robustness of the target drone at high angle of attack. ∞ The parameters of the control method are applied to the target drone's robust control system for high angle-of-attack flight to obtain the target drone's robustness H for high angle-of-attack flight. ∞ The controller output; the target drone's robustness to high angle-of-attack flight H ∞ The controller outputs the target drone's aileron deflection value, target drone elevator deflection value, and target drone rudder deflection value.

[0109] Specifically, step 5 includes:

[0110] Step 51: Set the control parameters. The objective of solving the system of linear matrix inequalities is: minTrace(W)

[0111] Wherein, min represents the minimization operation, and Trace(W) represents the trace of the parameter constraint matrix W;

[0112] Step 52: Use the Linear Matrix Inequality Toolbox (LMI) in Matlab simulation software to calculate the target drone's high angle-of-attack flight robustness H with minTrace(W) as the target. ∞ The control method uses a system of linear matrix inequalities for control parameters to obtain the target drone's robustness H at high angles of attack. ∞ The parameters of the control method are the positive definite matrix Y, the diagonal matrix Γ, the matrix V, and the constant τ;

[0113] Step 53: Set the target drone's high angle-of-attack flight robustness H ∞ The controller output u is:

[0114] Obtain the target drone aileron deflection λ ail Target drone elevator deflection λ ele Target drone rudder offset λ rud .

[0115] Step 6: Apply the target drone's aileron deflection value, elevator deflection value, and rudder deflection value to the target drone to enable the target drone to quickly, accurately, and simultaneously track high angle of attack commands, sideslip angle commands, and track roll angle commands; the high angle of attack command is a command with an angle of attack greater than 25°.

[0116] In summary, this invention provides a target drone with robust H-axis flight at high angles of attack. ∞ Control Method. In situations where the target drone exhibits model uncertainty, severe unsteadiness and nonlinearity, and is subject to external gust interference and severe coupling between yaw and roll directions, the proposed method utilizes accelerometers, gyroscopes, GPS, and airflow angle sensors to obtain the target drone's speed, angular velocity, roll angle, and airflow angle. These parameters are then used as feedback signals to calculate the aileron, elevator, and rudder deflections, enabling the following:

[0117] To ensure the asymptotic stability of the target drone's robust control system during high angle-of-attack flight;

[0118] The target drone can quickly and accurately track commands for high angle of attack, sideslip angle, and track roll angle.

[0119] The control method is robust to the target drone model parameters and does not require precise model parameters of the target drone. It only needs to have an estimate of the range of parameter variation to realize the target drone tracking control command even when the model parameters are uncertain.

[0120] The control method is robust to external gust interference, achieving a transfer function H of the robust control system for high angle-of-attack flight of the target drone from external gust interference. ∞ The norm is less than the index value, meaning that external gust interference has little impact on the robust control system of the target drone at high angle of attack.

[0121] The target drone's high angle-of-attack flight robust control system is quadratic D-stable, meaning that the closed-loop poles of the target drone's high angle-of-attack flight robust control system are within a preset limit domain.

[0122] The target drone's robust control system for high angle-of-attack flight meets the requirements for both steady-state and transient performance.

[0123] Example 2

[0124] This application provides a target drone with robust H-type high angle-of-attack flight. ∞ The control method is applied to a robust control system for high angle-of-attack flight. This embodiment, using a Lenovo Legion laptop, utilizes flight simulation software X-Plane and digital simulation software Matlab to test the robust H-axis flight performance of a target drone at high angle-of-attack. ∞ Experiments were conducted using the control method, yielding useful conclusions. The specific steps are as follows:

[0125] Step 1, see Figure 1 The target drone's velocity V and angular velocity ω are obtained from accelerometers, gyroscopes, GPS, and airflow angle sensors. b =[ω x ω y ω z ] T The target drone's roll angle μ and the target drone's airflow angle [αβ] T .

[0126] Step 2: Set the closed-loop pole limit region of the target drone's high angle-of-attack flight robust control system as follows:

[0127]

[0128] The closed-loop pole-restricted region is a circular region with its center at -12 and a half-value of 1.5.

[0129] Step 3: Set the tracking error model D and the target drone uncertainty structure for the target drone's high angle-of-attack flight robust control system;

[0130] For details, please refer to Figure 3 Step 3 includes:

[0131] Step 31: Set the desired angle of attack α for the target drone d Expected target drone sideslip angle β d Expected target drone roll angle μ d ;

[0132] Step 32: Set the airflow state error e1 for the target drone; the airflow state error e1 for the target drone is the desired angle of attack α of the target drone. d Expected target drone sideslip angle β d Expected target drone roll angle μ d The difference between the target drone's angle of attack α, sideslip angle β, and roll angle μ, and the specific form of the target drone's airflow state error e1, is: e1=[α d -αβ d -βμ d -μ] T ;

[0133] Step 33: Set the desired target angular velocity [ω] xd ω yd ω zd ] T ;

[0134] Step 34: Set the target drone angular velocity error e2; the target drone angular velocity error e2 is the desired target drone angular velocity [ω]. xd ω yd ω zd ] T With the target angular velocity [ω x ω y ω z ] T The difference between them, the specific form of the target angular velocity error e2 is: e2=[ω xd -ω x ω yd -ω y ω zd -ω z ] T ;

[0135] Step 34: Set the tracking error model D of the target drone's high angle-of-attack flight robust control system as follows:

[0136]

[0137] Where t is the simulation time;

[0138] Step 35: Set the target uncertainty structure for the target uncertainty ΔΦ and ΔΘ as follows:

[0139]

[0140] Step 4: Set the target drone's high angle-of-attack flight robustness H ∞ The control method comprises a set of linear matrix inequalities for control parameters; the set of linear matrix inequalities for control parameters consists of parameter-solved linear matrix inequalities and parameter-restricted linear matrix inequalities.

[0141] Specifically, specifically, see Figure 3 Step 4 includes:

[0142] Step 41: Set the target drone's high angle-of-attack flight robustness H ∞ The linear matrix inequality for solving the parameters of the control method is as follows:

[0143]

[0144] Wherein, the positive definite matrix Y, the diagonal matrix Γ, the matrix V, and the constant τ represent the target drone's robustness against high angle-of-attack flight, which is to be solved. ∞ The parameters of the control method are: I is the identity matrix; diagonal matrices D and E are the error weight matrix and input weight matrix of the target drone's robust control system for high angle-of-attack flight, respectively; and ξ is the H value of the external gust disturbance d to the target drone's robust control system for high angle-of-attack flight. ∞ Norm index; the H ∞ The norm index ξ is in the range of 0 < ξ < 0.5;

[0145] Step 42: Set the target drone's high angle-of-attack flight robustness H ∞ The parameter constraint linear matrix inequality for the control method is:

[0146]

[0147] Wherein, W is the parameter constraint matrix, and I is the identity matrix;

[0148] Step 43: Solve the linear matrix inequalities with parameters and parameter constraints to obtain the target drone's high angle-of-attack flight robustness H. ∞ The control parameters of the control method are a set of linear matrix inequalities.

[0149] Step 5: Set the objective of solving the system of linear matrix inequalities for control parameters. Use the Linear Matrix Inequality Toolbox (LMI) in Matlab simulation software to calculate the objective and obtain the target drone's robustness H at high angle of attack. ∞ Parameters of the control method; robustness of the target drone at high angle of attack. ∞ The parameters of the control method are applied to the target drone's robust control system for high angle-of-attack flight to obtain the target drone's robustness H for high angle-of-attack flight. ∞ The controller output; the target drone's robustness to high angle-of-attack flight H ∞ The controller outputs the target drone's aileron deflection value, target drone elevator deflection value, and target drone rudder deflection value.

[0150] For details, please refer to Figure 3 Step 5 includes:

[0151] Step 51: Set the control parameters. The objective of solving the system of linear matrix inequalities is: minTrace(W)

[0152] Wherein, min represents the minimization operation, and Trace(W) represents the trace of the parameter constraint matrix W;

[0153] Step 52: Use the Linear Matrix Inequality Toolbox (LMI) in Matlab simulation software to calculate the target drone's high angle-of-attack flight robustness H with minTrace(W) as the target. ∞ The control parameter linear matrix inequalities of the control method are used to obtain the target drone's robustness H at high angle of attack. ∞ The control method comprises a positive definite matrix Y, a diagonal matrix Γ, a matrix V, and a constant τ; the target drone's high angle-of-attack flight robustness H... ∞ The value obtained by solving the positive definite matrix Y of the control method is:

[0154]

[0155] The target drone is robust to high angle-of-attack flight. ∞ The parameter matrix V obtained by solving the control method is:

[0156]

[0157] Step 53: Set the target drone's high angle-of-attack flight robustness H ∞ The controller output u is:

[0158] Obtain the target drone aileron deflection λ ail Target drone elevator deflection λ ele Target drone rudder offset λ rud .

[0159] Step 6: Apply the target drone's aileron deflection value, elevator deflection value, and rudder deflection value to the target drone to enable the target drone to quickly, accurately, and simultaneously track high angle of attack commands, sideslip angle commands, and track roll angle commands.

[0160] Simulation results based on X-Plane and Matlab: Figure 4 The target drone's response curve for tracking angle-of-attack commands is shown in the diagram. The red dashed line represents the target drone's desired angle-of-attack command, and the blue solid line represents the target drone's actual angle of attack. It can be seen that the target drone's robustness to high angle-of-attack flight H provided in this application is... ∞ Under the control method, the target drone can overcome the uncertainty of the model and the nonlinearity and unsteadiness of the aerodynamic parameters caused by large angle of attack, and achieve rapid, overshoot-free, and steady-state error-free tracking of angle-of-attack step commands, including 40° large angle-of-attack commands. Figure 5 This is a target drone tracking sideslip angle command response curve, where the red dashed line represents the target drone's desired sideslip angle command, and the blue solid line represents the target drone's actual sideslip angle; Figure 5This is a target drone's response curve to a roll angle command tracking trajectory. The red dashed line represents the target drone's desired roll angle command, and the blue solid line represents the target drone's actual roll angle. Figure 5 and Figure 6 It can be seen that the target drone provided in this application is robust to high angle-of-attack flight. ∞ Under the control method, the target drone can overcome the high coupling between yaw and roll directions, and achieve tracking of ramp sideslip angle and track roll angle commands with a maximum amplitude of 5° under different angles of attack, including a large angle of attack of 40°. This invention has strong engineering application value.

Claims

1. A target drone with robust high angle-of-attack flight The control method is characterized by, The method is applied to a robust control system for high angle-of-attack flight of a target drone, and the method includes: Step 1: Obtain the target drone's speed, angular velocity, roll angle, and airflow angle using an accelerometer, gyroscope, GPS, and airflow angle sensor; the airflow angle consists of the target drone's angle of attack and sideslip angle. Step 2: Set the limiting domain of the closed-loop poles of the target drone's high angle-of-attack flight robust control system; Step 3: Set up the tracking error model and target uncertainty structure for the target drone's robust control system for high angle of attack flight; Step 4: Set the target drone for high angle of attack flight robustness The control method comprises a set of linear matrix inequalities for control parameters; the set of linear matrix inequalities for control parameters consists of parameter-solved linear matrix inequalities and parameter-restricted linear matrix inequalities. Step 5: Set the objective of solving the system of linear matrix inequalities using control parameters. Linear Matrix Inequality Toolbox in Simulation Software The target objective is calculated to obtain the target drone's robustness at high angles of attack. Control method parameters; robustness of target drone flight at high angles of attack. The parameters of the control method are applied to the target drone's robust control system for high angle-of-attack flight to obtain the target drone's robustness for high angle-of-attack flight. The controller output; the target drone is robust to high angle-of-attack flight. The controller outputs the target drone's aileron deflection value, target drone elevator deflection value, and target drone rudder deflection value. Step 6: Apply the target drone's aileron deflection, elevator deflection, and rudder deflection to the target drone to enable the target drone to quickly and accurately track high angle-of-attack commands, sideslip angle commands, and track roll angle commands simultaneously; the high angle-of-attack command is an angle of attack greater than... The instructions.

2. The method according to claim 1, characterized in that, Step 1 includes: Step 11: Obtain the target drone velocity in the airflow coordinate system using the accelerometer. The target drone's airflow coordinate system has , , Three-axis , , The three axes point forward, right, and downward, respectively; the target speed... The direction is Axial direction; Step 12: Obtain the target drone's angle of attack using the airflow angle sensor. and target drone sideslip angle ; Step 13: Obtain the angular velocity of the target drone in the target drone body coordinate system using the gyroscope. The target body coordinate system has , , Three-axis , , The three axes point forward, right, and downward, respectively; , and In the target drone body coordinate system Angular velocity of the target drone in the axial direction Angular velocity of the target drone in the axial direction and Angular velocity of the target drone in the axial direction; Step 14: Obtain the target drone's roll angle from GPS. .

3. The method according to claim 1, characterized in that, Step 2 includes: Step 21: Set the closed-loop pole constraint region of the target drone's high angle-of-attack flight robust control system to a circular region located to the left of the imaginary axis of the Laplace-varying complex plane; the closed-loop pole constraint region The specific form is: ； Among them, the , , Let be the location of the center of the circular region. Let be the radius of the circular region. For the complex field of Laplace transform, yes The conjugate of complex numbers.

4. The method according to claim 1, characterized in that, Step 3 includes: Step 31: Set the desired angle of attack for the target drone Expected target drone sideslip angle Expected target drone roll angle ; Step 32: Set the airflow state error for the target drone The target airflow state error To the desired angle of attack of the target drone Expected target drone sideslip angle Expected target drone roll angle Angle of attack of the target drone Target drone sideslip angle target drone trajectory roll angle The difference between them, the airflow error of the target drone The specific form is: ; Step 33: Set the desired target angular velocity ; Step 34: Set the target drone angular velocity error The target angular velocity error Desired target angular velocity Angular velocity of the target The difference between them, the target drone angular velocity error The specific form is: ； Step 35: Set up the tracking error model for the target drone's robust control system for high angle-of-attack flight. The tracking error model of the target drone's high angle-of-attack flight robust control system. The specific form is: ； Among them, the The total tracking error of the target drone. Let be the derivative of the total tracking error of the target drone with respect to time. To match the speed of the target drone The relevant target state matrix, To match the speed of the target drone The relevant target input matrix, and For the target machine uncertainty, This indicates external gust interference with the target drone. This indicates that the target drone is robust during high angle-of-attack flight. The controller output includes the target drone's aileron deflection. Target drone elevator deflection rudder offset of the target drone ; Step 36: Set the target uncertainty and The target uncertainty structure is as follows: ； Among them, matrix , and The dimension is The known constant matrix represents the uncertainties. and Structural information; It is a time-varying unknown matrix, representing uncertainties. and Uncertainty; It is a diagonal matrix. It is an identity matrix; wherein, the matrix , and The value of should satisfy the following: .

5. The method according to claim 1, characterized in that, Step 4 includes: Step 41: Set the target drone for high angle of attack flight robustness The linear matrix inequality for solving the parameters of the control method is as follows: ； Among them, positive definite matrix diagonal matrix ,matrix and constant Robustness to high angle-of-attack flight of the target drone to be solved Parameters of the control method The identity matrix and the diagonal matrix and diagonal matrix These are the error weight matrix and input weight matrix of the target drone's robust control system for high angle-of-attack flight. External gusts of wind Robust control system for target drone high angle of attack flight Norm index; the Norm index The range is ; Step 42: Set the target drone for high angle of attack flight robustness The parameter constraint linear matrix inequality for the control method is: ； Among them, the For parameter constraint matrix, It is the identity matrix; Step 43: Solve the linear matrix inequalities with parameters and parameter constraints to obtain the target drone's robustness at high angle of attack. The control parameters of the control method are a set of linear matrix inequalities.

6. The method according to claim 1, characterized in that, Step 5 includes: Step 51: Set the control parameters and the objective of solving the system of linear matrix inequalities as follows: Among them, the This represents the minimization operation. Represents the parameter constraint matrix traces; Step 52: Use Linear Matrix Inequality Toolbox in Simulation Software by Robustness of target drone flight at high angle of attack for target calculation The system of linear matrix inequalities for the control parameters of the control method is used to obtain the target drone's robustness at high angles of attack. Positive definite matrix of parameters of the control method diagonal matrix ,matrix and constant ; Step 53: Set the target drone for high angle of attack flight robustness Controller output for: ； Obtain the target drone aileron deflection value Target drone elevator deflection rudder offset of the target drone .

7. The method according to claim 3, characterized in that, The closed-loop pole-limited region of the target drone's high angle-of-attack flight robust control system The structure is ;in, It refers to the complex field of the Laplace transform. yes The conjugate complex number, symmetric matrix sum matrix Used to describe the shape of the closed-loop pole confinement region; The closed-loop pole limiting region of the target drone's high angle-of-attack flight robust control system can ensure that the target drone's high angle-of-attack flight robust control system meets the requirements of steady-state performance and transient performance. Set the closed-loop pole constraint region to a circular region, and set the matrix... ,matrix ; wherein, the , , Let be the location of the center of the circular region. Let be the radius of the circular region; If the robust control system for high angle-of-attack flight of the target drone requires fast convergence speed and small steady-state error, then the parameters... The scope is: ,parameter The range of values ​​for is: ; The robust control system for target drones flying at high angles of attack requires small overshoot and high climb rate, therefore the parameters... The scope is: ,parameter The range of values ​​for is: .

8. A robust control system for high angle-of-attack flight of a target drone, characterized in that, The target drone described in any one of claims 1 to 7 is robust to high angle-of-attack flight. The control method is deployed on the target drone's high angle-of-attack flight robust control system, which includes an accelerometer, gyroscope, GPS, airflow angle sensor, and target drone high angle-of-attack flight robust control system. Controller, ailerons, elevator, and rudder; where: the target drone's velocity is measured by the accelerometer. Angular velocity of the target drone measured by a gyroscope The roll angle of the target drone's trajectory measured by GPS Target angle of attack measured by airflow angle sensor and target drone sideslip angle The target drone is robust to high angle-of-attack flight. Controller inputs; robust target drone flight at high angles of attack The controller outputs the deflection values ​​for the ailerons, elevator, and rudder.

Images