Face symmetry carrier rocket self-adaptive control method based on double-ring finite time convergence

Abstract

The invention discloses a double-ring finite time convergence plane symmetry carrier rocket adaptive control method, which comprises the following steps of: firstly, designing a high-performance attitude controller by depending on a fixed time sliding mode control theory, and realizing accurate attitude adaptive tracking by combining with an adaptive disturbance observer; then, a position loop control scheme is constructed based on the finite time theory, a six-degree-of-freedom coupling control framework is constructed through finite time collaborative design of a position loop and an attitude loop, the dynamic response characteristics of the position and the attitude are optimized, and the system stability in the fast maneuvering process is ensured; and finally, under the conditions of strong aerodynamic interference and uncertain parameters, for the designed double-loop finite time controller, systematic examination of reliability and control performance is carried out, and robustness and superiority of the double-loop finite time controller under the conditions of complex external interference and uncertain parameters are verified. According to the method, the high-precision attitude tracking control problem of the plane-symmetric rocket can be solved, the response performance of a rocket multivariable system is improved, and the flight reliability is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic control, in particular to a double-loop finite-time converging face-symmetrical launch vehicle adaptive control method. BACKGROUND

[0002] With the increasing complexity of space missions, large launch vehicles are facing stringent control requirements such as rapid attitude maneuvering, accurate pointing, and strong anti-interference. Traditional asymptotic convergence control methods have inherent limitations in convergence speed, large-angle singularity, and anti-interference ability, making it difficult to meet the high-performance control requirements of new-generation launch vehicles. Therefore, the designed control method needs to have rapid convergence and anti-interference ability. Since face-symmetrical launch vehicles have only been studied as a new type of launch vehicle in recent years, there are relatively few related literature. Domestic and foreign scholars have made a series of important progress in the fields of fixed-time convergence control, preset performance control, and spacecraft attitude control, but there is still room for further exploration in adaptive fixed-time cooperative control in face-symmetrical rocket large-angle maneuvering control, robustness enhancement in strong disturbance environment, and real-time optimization of control algorithm.

[0003] Currently, the research on finite-time control theory has made some important progress. In single-input single-output systems, based on finite-time convergence theory, researchers have proposed various improved algorithms such as finite-time sliding mode control and adaptive finite-time control to solve the problem of rapid convergence of nonlinear systems under parameter uncertainty and external disturbance. However, existing researches are mostly focused on relatively simple single-input single-output systems, and their control structure and stability analysis methods are difficult to directly apply to complex multi-input multi-output systems such as reusable launch vehicles. In particular, the attitude control of launch vehicles has the following special challenges: 1) strong coupling characteristics caused by six degrees of freedom motion; 2) nonlinear changes of aerodynamic force / momentum with flight state. Therefore, it is necessary to develop a finite-time controller suitable for multi-input multi-output systems for the multi-variable characteristics of launch vehicles to ensure that the three-axis attitude angles of the rocket are accurately tracked within a finite time while meeting the dynamic coordination requirements between channels. SUMMARY

[0004] The present application provides a double-loop finite-time converging face-symmetrical launch vehicle adaptive control method, which can handle the rapid maneuvering and large-angle attitude tracking requirements of large face-symmetrical general core launch vehicles, improve the accuracy of the attitude control system, and improve the success rate of rocket soft landing. It is closer to actual engineering application while improving performance.

[0005] The present application provides a double-loop finite-time converging face-symmetrical launch vehicle adaptive control method, which includes the following steps:

[0006] Step 1, considering the uncertainty and internal and external interference, an attitude control model of a face-symmetrical launch vehicle in a dynamic landing stage is established;

[0007] Step 2, an adaptive disturbance observer is designed to estimate the compound disturbance composed of structural disturbance, wind disturbance and engine vibration disturbance in real time, so that the adaptive estimation of parameters is realized;

[0008] Step 3, based on the fixed-time sliding mode control theory, a high-performance attitude controller is designed in combination with the adaptive disturbance observer, so that precise high-speed large attitude maneuver tracking is realized;

[0009] Step 4, in the landing coordinate system, a rigid body motion dynamics model is constructed according to the rigid body model center of mass translational dynamics equation of the rocket in the dynamic landing stage;

[0010] Step 5, a position loop control scheme is constructed based on the finite time theory, and a six-degree-of-freedom coupling control framework is constructed through the finite time collaborative design of the position loop and the attitude loop;

[0011] Step 6, based on the six-degree-of-freedom coupling control framework, a parameter setting law involved in the control law is given;

[0012] Step 7, considering the actual constraint conditions, the adaptive control algorithm of the face-symmetrical launch vehicle is verified.

[0013] Optionally, in an embodiment of the present application, step 1 specifically comprises:

[0014] The attitude control model of the face-symmetrical launch vehicle in the dynamic landing stage is:

[0015]

[0016] Wherein, is the roll angle, yaw angle and pitch angle in the landing coordinate system; ω=[ω x ω y ω z ] T is the three-axis attitude angular velocity in the rocket body coordinate system; is the equivalent swing angle command of the rocket pitch, yaw and roll channels, and ΔD is the total disturbance, including structural disturbance, wind disturbance and engine vibration disturbance; the matrices R, Ω and E are respectively:

[0017]

[0018] Wherein, and are the structural disturbance moments; ɑ w and β w are the additional wind attack angle and wind sideslip angle respectively:

[0019]

[0020] where W is the wind speed; A w is the wind direction angle; A0 is the launch azimuth angle; V is the mass center velocity; Θ is the velocity inclination angle;

[0021] Take the state vector as Control variable where γ c is the reference roll angle, ψ c is the reference yaw angle, is the reference yaw angle, is the first derivative of the roll angle, is the first derivative of the reference roll angle, is the first derivative of the yaw angle, is the first derivative of the reference yaw angle, is the first derivative of the pitch angle, is the first derivative of the reference pitch angle, the attitude control model is converted into the following form:

[0022]

[0023] where, is the second derivative of the reference attitude angle, g(x) = R·E, d(t) = R·ΔD, and it is assumed that d(t) is bounded, that is, there is a constant δ>0 such that |d(t)|≤δ holds for all t≥0.

[0024] Optionally, in an embodiment of the present application, in step 2, the designed adaptive disturbance observer is:

[0025]

[0026] where, is the disturbance estimation value, l(x) is the observer gain matrix, l(x) = [l1l2l3] T , l1, l2, l3>0, is the designed auxiliary function;

[0027] The disturbance estimation value based on the output of the adaptive disturbance observer The time-varying control parameter k(t) is adaptively adjusted, and the control parameter k(t) is given in the following form:

[0028]

[0029] where k0>0 is the reference gain, and λ2>0 is the controller parameter.

[0030] Optionally, in an embodiment of the present application, in step 3, based on the fixed-time sliding mode control theory, the following sliding surface s(x) is designed for the conversion model:

[0031]

[0032] wherein the sliding surface parameters λ1>0, μ1>0, β>0, so that has the characteristics of terminal sliding mode, that is, the sliding surface guarantees that the system states x1and x2converge to zero in a fixed time;

[0033] The saturation function expression is as follows:

[0034]

[0035] The designed controller is:

[0036]

[0037] wherein, μ2>0 is a controller parameter, and λ2is a controller parameter, which are adaptively adjusted by step 2;

[0038] -g -1 (x)f(x) belongs to the error dynamic compensation term, which is used to compensate the known dynamics of the system;

[0039] is a nonlinear reaching law, wherein, guarantees the reaching speed; the exponential term makes the reaching law exhibit different convergence characteristics at different stages, that is, when far away from the sliding surface, the exponential is large, and when close to the sliding surface, the exponential gradually decreases;

[0040] is an additional nonlinear compensation term;

[0041] The stable time satisfies:

[0042]

[0043] wherein, is a stable time calculation parameter.

[0044] Optionally, in an embodiment of the present application, in step 4, the dynamics model of the rigid body motion is:

[0045]

[0046] wherein, r=[r x r y r z ] TThe position vector r of the rocket mass center relative to the landing coordinate system is decomposed along the landing coordinate system O-x4y4z4 to obtain; the velocity vector v of the rocket mass center relative to the landing coordinate system is decomposed along the landing coordinate system O-x4y4z4 to obtain v=[v x v y v z ] T ; g is the gravitational acceleration vector, which is represented as in the landing coordinate system:

[0047]

[0048] wherein g r is the component of the earth gravitational acceleration in the geocentric radius; g ωe is the component of the earth gravitational acceleration in the direction of the earth rotation; (x, y, z) are the coordinates of the rocket in the landing coordinate system; R 0x , R 0y , R 0z are the components of the geocentric radius of the origin of the landing coordinate system in the landing coordinate system; ω ex , ω ey , ω ez are the components of the earth rotation angular velocity in the landing coordinate system;

[0049] T is the thrust of the rocket along the rocket body axis in the rocket body coordinate system, and the thrust vector T is represented as in the landing coordinate system:

[0050]

[0051] wherein, is the rotation matrix from the landing coordinate system to the rocket body coordinate system;

[0052] a D is the air resistance acceleration vector of the rocket in the landing coordinate system:

[0053]

[0054] wherein C D is the aerodynamic axial force coefficient; S ref is the aerodynamic reference area of the rocket; and ρ is the atmospheric density.

[0055] Optionally, in an embodiment of the present application, in step 5, a position loop control scheme is constructed based on the finite time theory, and the sliding mode surface is defined as:

[0056] s=e v +λe

[0057] wherein e=r-r d , are the position error and the velocity error, respectively, and r ddesired position, and v d desired velocity, λ>0;

[0058] The control law is designed as:

[0059] u = -k1|s| α ·sat(s)-k2s

[0060] where k1>0, k2>0;

[0061] The system stability time satisfies:

[0062]

[0063] where c=k1·2 (1+α) / 2 ,

[0064] Optionally, in an embodiment of the application, in step 6, the parameter setting law involved in the attitude loop control law is:

[0065] The feedback linear compensation term is -g -1 (x)f(x) is a dynamic compensation term known to the system and does not involve parameter adjustment; the nonlinear approach law part is where k satisfies δ is the upper bound of the system disturbance, which is adaptively adjusted by the disturbance observer; λ2 controls the steepness of the exponential function, and the greater it is, the faster the convergence when far away from the sliding mode surface; μ2 adjusts the smoothness of the exponential, and increasing it reduces the chattering when close to the sliding mode surface; is an adaptive exponential term that automatically adjusts the strength of the control input change with the error size; the nonlinear compensation part is where λ1 controls the nonlinear strength of the sliding mode surface, and increasing it enhances the control ability near the sliding mode surface; μ1 weakens the compensation term explosion in extreme cases; β adjusts the overall strength of the nonlinear compensation part, from 0.1 to 2;

[0066] The parameter setting law involved in the position loop control law is: λ determines the weight of the position error and the speed error in the error dynamics, controls the response speed of the sliding mode surface, and is set to 1-10; α is the nonlinear power term, which determines the convergence speed and controls the chattering to a certain extent, and is set to 0-1; k1 is the nonlinear approach term gain, which mainly controls the convergence speed, and the greater the value, the greater the approach speed, and the control input amplitude also increases; k2 is the linear feedback gain; in the parameter setting process, the following order is followed: first, fix α and λ, and increase k1 until the desired convergence speed is met, and then adjust k2 to suppress overshoot and oscillation.

[0067] Optionally, in one embodiment of the present application, step 7 specifically comprises: considering the actual constraint conditions of strong aerodynamic interference and parameter uncertainty, verifying the robustness and superiority of the adaptive control algorithm of the face-to-face symmetric launch vehicle under complex external interference and parameter uncertainty.

[0068] The double-loop finite-time converging face-to-face symmetric launch vehicle adaptive control method of the embodiment of the present application has the following beneficial effects:

[0069] 1. The present application breaks through the limitations of traditional single-input single-output finite-time control methods and innovatively designs a multi-input multi-output finite-time control architecture for a six-degree-of-freedom rocket model.

[0070] 2. The present application integrates adaptive mechanisms and finite-time control theory, enhances system robustness while ensuring convergence speed, and experiments show that the control performance is good under ±20% parameter perturbation and wind field interference; compared with traditional PID control, the coupling effect suppression effect is also significantly improved.

[0071] Additional aspects and advantages of the present application will be partially given in the following description, partially will become obvious from the following description, or will be understood through the practice of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0072] The above and / or additional aspects and advantages of the present application will become apparent and more readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:

[0073] Figure 1 A flowchart of a double-loop finite-time converging face-to-face symmetric launch vehicle adaptive control method according to an embodiment of the present application is shown in Figure 1;

[0074] Figure 2 A control structure diagram of the double-loop finite-time converging adaptive control method for face-to-face symmetric launch vehicles of the present application is shown in Figure 2;

[0075] Figure 3 An attitude angle curve of a face-to-face symmetric launch vehicle in the powered landing phase under the conditions of disturbance, structural disturbance, engine buffeting and position ring wind resistance is shown in Figure 3;

[0076] Figure 4 An attitude angle tracking error curve of a face-to-face symmetric launch vehicle in the powered landing phase under the conditions of disturbance, structural disturbance, engine buffeting and position ring wind resistance is shown in Figure 4;

[0077] Figure 5 An attitude angle rate curve of a face-to-face symmetric launch vehicle in the powered landing phase under the conditions of disturbance, structural disturbance, engine buffeting and position ring wind resistance is shown in Figure 5;

[0078] Figure 6is the three-axis position curve of the face-symmetrical launch vehicle in the powered landing phase considering the interference, structural interference, engine chattering and position loop wind resistance conditions;

[0079] Figure 7 is the three-axis velocity curve of the face-symmetrical launch vehicle in the powered landing phase considering the interference, structural interference, engine chattering and position loop wind resistance conditions;

[0080] Figure 8 is the roll angle Monte Carlo simulation curve of the face-symmetrical launch vehicle in the powered landing phase under the condition of ±20% parameter uncertainty;

[0081] Figure 9 is the pitch angle Monte Carlo simulation curve of the face-symmetrical launch vehicle in the powered landing phase under the condition of ±20% parameter uncertainty;

[0082] Figure 10 is the yaw angle Monte Carlo simulation curve of the face-symmetrical launch vehicle in the powered landing phase under the condition of ±20% parameter uncertainty;

[0083] Figure 11 is the rocket attitude angle comparison curve under the conditions of structural interference, wind interference and engine chattering interference, and the position loop control algorithm respectively adopts finite time sliding mode, PD control and quartic polynomial guidance;

[0084] Figure 12 is the rocket attitude angle rate comparison curve under the conditions of structural interference, wind interference and engine chattering interference, and the position loop control algorithm respectively adopts finite time sliding mode, PD control and quartic polynomial guidance;

[0085] Figure 13 is the core engine swing angle command comparison curve under the conditions of structural interference, wind interference and engine chattering interference, and the position loop control algorithm respectively adopts finite time sliding mode, PD control and quartic polynomial guidance;

[0086] Figure 14 is the boost engine swing angle command comparison curve under the conditions of structural interference, wind interference and engine chattering interference, and the position loop control algorithm respectively adopts finite time sliding mode, PD control and quartic polynomial guidance;

[0087] Figure 15 is the rocket horizontal position comparison curve under the conditions of structural interference, wind interference and engine chattering interference, and the position loop control algorithm respectively adopts finite time sliding mode, PD control and quartic polynomial guidance;

[0088] Figure 16 is the rocket lateral position comparison curve under the conditions of structural interference, wind interference and engine chattering interference, and the position loop control algorithm respectively adopts finite time sliding mode, PD control and quartic polynomial guidance;

[0089] Figure 17 is the rocket vertical position comparison curve under the condition of structural disturbance, wind disturbance and engine chattering disturbance, and the position loop control algorithm respectively adopts finite time sliding mode, PD control and quartic polynomial guidance;

[0090] Figure 18 is the rocket three-axis thrust comparison curve under the condition of structural disturbance, wind disturbance and engine chattering disturbance, and the position loop control algorithm respectively adopts finite time sliding mode, PD control and quartic polynomial guidance. DETAILED DESCRIPTION

[0091] Embodiments of the present application will be described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by reference to the accompanying drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application.

[0092] As shown in Figure 1 and Figure 2 , the double-loop finite time convergent face-symmetric launch vehicle adaptive control method comprises the following steps:

[0093] Step 1, considering the uncertainty and internal and external disturbances, an attitude control model of the face-symmetric launch vehicle in the dynamic landing stage is established.

[0094] The attitude control model of the face-symmetric launch vehicle in the dynamic landing stage is established as follows:

[0095]

[0096] wherein, is the roll angle, yaw angle and pitch angle in the landing coordinate system; ω=[ω x ω y ω z ] T is the three-axis attitude angular velocity in the rocket body coordinate system; is the equivalent swing angle command of the rocket pitch, yaw and roll channels, ΔD is the total disturbance, including structural disturbance, wind disturbance and engine vibration disturbance; matrices R, Ω and E are respectively:

[0097]

[0098] wherein, and are the structural disturbance moments; α w and β w are the additional wind attack angle and wind sideslip angle respectively:

[0099]

[0100] where: W is wind speed; A w is wind direction angle; A0 is launch azimuth angle; V is mass center velocity; Θ is velocity inclination angle.

[0101] Take state vector as Control variable where, γ c is reference roll angle, ψ c is reference yaw angle, is reference yaw angle, is roll angle first derivative, is reference roll angle first derivative, is yaw angle first derivative, is reference yaw angle first derivative, is pitch angle first derivative, is reference pitch angle first derivative, the model is converted to the following form:

[0102]

[0103] where, g(x)=R·E,d(t)=R·ΔD,ΔD is total disturbance, including structural disturbance, wind disturbance, engine vibration disturbance, and assume that d(t) is bounded, that is, there is a constant δ>0 such that |d(t)|≤δ for all t≥0.

[0104] Step 2, design an adaptive disturbance observer, estimate the composite disturbance composed of structural disturbance, wind disturbance and engine vibration disturbance in real time, to realize adaptive estimation of parameters.

[0105] The designed adaptive disturbance observer is:

[0106]

[0107] where, g(x)=R·E,l(x) is observer gain matrix l(x)=[l1l2l3] T ,l1,l2,l3>0, is the designed auxiliary function.

[0108] The disturbance estimation value based on the output of the observer Adaptive adjustment of time-varying control parameter k(t), specifically, the control parameter k(t) is given by the following form:

[0109]

[0110] where, k0>0 is the reference gain, λ2>0 is the controller parameter.

[0111] Step 3: Based on the fixed-time sliding mode control theory, a high-performance attitude controller is designed by combining an adaptive disturbance observer to achieve precise high-speed large attitude maneuver tracking.

[0112] Based on the rigid body attitude control model of the rocket-powered landing phase, the attitude angle and attitude angular velocity of the face-symmetrical launch vehicle are taken as the outputs of the system, and the engine swing angle of the pitch, yaw, and roll channels is taken as the input of the control system.

[0113] Based on the fixed-time sliding mode control theory, the following sliding mode surface s(x) is designed for the conversion model:

[0114]

[0115] where: sliding mode surface parameters λ1>0, μ1>0, β>0, so that has the characteristics of terminal sliding mode, that is, the sliding mode surface can guarantee that the system states x1 and x2 converge to zero in a fixed time. The saturation function expression is as follows:

[0116]

[0117] The controller is designed as:

[0118]

[0119] where: μ2>0 is the controller parameter, and λ2 is the controller parameter, which is adjusted adaptively by step 2.

[0120] where, -g -1 (x)f(x) belongs to the error dynamic compensation term, which is used to compensate for the known dynamics of the system, similar to the feedback linearization method, so that the closed-loop system is equivalent to a standard sliding mode control system.

[0121] is a nonlinear reaching law. The nonlinear index design of the reaching law enables the control input to be adaptively adjusted under different states, effectively improving the convergence performance and suppressing chattering. Where, guarantees the reaching speed; the exponential term makes the reaching law exhibit different convergence characteristics at different stages: when far from the sliding mode surface, the index is larger, accelerating the system to reach the sliding mode surface; when close to the sliding mode surface, the index gradually decreases, reducing the reaching chattering and improving the steady-state accuracy.

[0122] It can be understood as an additional nonlinear compensation term, which is derived from the nonlinear term of the sliding mode surface, which ensures that the controller can adapt to different states during the process of the sliding mode surface tending to zero, and improves the stability of the system. The introduction of the ln|x1| term makes the compensation term effective when the absolute value of |x1| is small, and enhances the convergence of the low-speed section of the system.

[0123] The stable time satisfies:

[0124]

[0125] Wherein: is a stable time calculation parameter.

[0126] Step 4, in the landing coordinate system, according to the rigid body model of the rocket powered landing segment, the center of mass translational dynamics equation is established, and the dynamics model of rigid body motion is constructed.

[0127] Based on the center of mass translational equation in the landing coordinate system, the rocket dynamics model is established, which lays the foundation for the position ring controller design of the face-symmetrical launch vehicle.

[0128] In the landing coordinate system, according to the rigid body model of the rocket powered landing segment, the center of mass translational dynamics equation is established, and the dynamics model of rigid body motion is constructed.

[0129]

[0130] Wherein, r = [r x r y r z ] T is the position vector of the rocket center of mass relative to the landing coordinate system, which can be obtained by decomposing along the landing coordinate system O-x4y4z4; the velocity vector v of the rocket center of mass relative to the landing coordinate system can be obtained by decomposing along the landing coordinate system O-x4y4z4v = [v x v y v z ] T ; g is the acceleration vector of gravity, which can be expressed in the landing coordinate system as:

[0131]

[0132] Wherein, g r is the component of the earth's gravitational acceleration in the geocentric vector; g ωe is the component of the earth's gravitational acceleration in the direction of the earth's rotation; (x, y, z) are the coordinates of the rocket in the landing coordinate system; R 0x , R 0y , R 0z are the components of the geocentric vector at the origin of the landing coordinate system in the landing coordinate system; ω ex , ωey , ω ez are the components of the earth rotation angular velocity in the landing coordinate system respectively.

[0133] T is the rocket thrust along the rocket body axis in the rocket body coordinate system. The thrust vector T can be expressed in the landing coordinate system as:

[0134]

[0135] where, is the rotation matrix from the landing coordinate system to the rocket body coordinate system; a D is the rocket air resistance acceleration vector in the landing coordinate system:

[0136]

[0137] where, C D is the aerodynamic axial force coefficient; S ref is the rocket aerodynamic reference area; and p is the atmospheric density.

[0138] Step 5, based on the finite time theory, a position loop control scheme is constructed. Through the finite time cooperative design of the position loop and the attitude loop, a six-degree-of-freedom coupled control framework is constructed.

[0139] Based on the rigid body kinematics model of the rocket-powered landing phase, the expected attitude angle obtained by solving the face-symmetrical launch vehicle is taken as the output of the system, and the expected three-axis position and three-axis velocity are taken as the input of the control system.

[0140] Based on the finite time theory, a position loop control scheme is constructed. The sliding surface is defined as:

[0141] s = e v + λe

[0142] where, e = r - r d , are the position error and velocity error respectively, r d is the expected position, and v d is the expected velocity, and λ > 0.

[0143] The control law is designed as:

[0144] u = -k1|s| α · sat(s) - k2s

[0145] where, k1 > 0, k2 > 0, and 0 < a < 1.

[0146] The system stable time satisfies:

[0147]

[0148] where c = k1-2 (1+α) / 2 , 0 < γ < 1.

[0149] Step 6, based on the six-degree-of-freedom controller design, the parameter tuning law involved in the control law is given.

[0150] The parameter tuning law involved in the attitude loop control law is: the feedback linear compensation term -g -1 (x)f(x) is the known dynamic compensation term of the system, which does not involve parameter tuning. The nonlinear reaching law part is where k must satisfy (δ is the upper bound of system disturbance), which is adaptively adjusted by the disturbance observer; λ2 controls the steepness of the exponential function, the greater the steeper the convergence when far from the sliding surface; μ2 adjusts the smoothness of the exponential, which can reduce the chattering when close to the sliding surface; is the adaptive exponential term, which can automatically adjust the strength of the control input as the error changes. The nonlinear compensation part is where λ1 controls the nonlinear strength of the sliding surface, which can enhance the control ability near the sliding surface; μ1 can weaken the explosion of the compensation term in extreme cases; β can adjust the overall strength of the nonlinear compensation part, which can be adjusted from 0.1 to 2.

[0151] The parameter tuning law involved in the position loop control law is: λ determines the weight of the position error and the speed error in the error dynamic, which can control the response speed of the sliding surface, generally set to 1-10; α is the nonlinear power term, which can determine the convergence speed and control the chattering to a certain extent, generally set to 0-1; k1 is the nonlinear reaching term gain, which mainly controls the convergence speed, in general, the larger the value, the faster the reaching speed, but the control input amplitude also increases; k2 is the linear feedback gain, which can improve the robustness of the system, enhance the linear damping, and improve the stability and disturbance suppression ability. In general, in the parameter tuning process, the following order can be followed: first, fix α and λ, increase k1 until the desired convergence speed is met, and then adjust k2 to suppress overshoot and oscillation.

[0152] Step 7, considering the actual constraints such as strong aerodynamic disturbance and parameter uncertainty, verify the robustness and superiority of the control algorithm under complex external disturbance and parameter uncertainty.

[0153] Verify the anti-interference ability of the system under the double-loop finite time control, add wind disturbance, structural disturbance and engine chattering in the rocket attitude model, and add parameter uncertainty according to the analysis of model force and torque. And compared with the traditional algorithm.

[0154] where, represents the uncertainty of the aerodynamic coefficient, d0 represents the uncertainty of the moment of inertia, Represents structural disturbances and moment uncertainties. The coefficients of the pitch axis centroid dynamic model represent the unknown angle of attack.

[0155] The effectiveness of this invention is verified through simulation below. The simulation parameters are as follows:

[0156] Initial attitude angle vector Initial position vector [r x0 ,r y0 ,r z0 ] T =[50,15,1000] T m, initial velocity vector [v] x0 ,v y0 ,v z0 ] T =[-6,2,-40] T m / s, target position vector [r x ,r y ,r z ] T =[0,0,0] T m, target velocity vector [v] x ,v y ,v z ] T =[0,0,0] T m / s, target acceleration vector [a x ,a y ,a z ] T =[-0.1,0,9.8] T m 2 / s. The simulation time is set to 50s. The launch azimuth angle A0 = 3.368rad, and the rocket aerodynamic axial coefficient C D =1, rocket aerodynamic reference area S ref =10m 2 The total mass m = 180 kg, and the atmospheric density ρ = 1.225 kg / m³. 2 The acceleration due to gravity is g = 9.8 m / s². 2 Maximum thrust 2000N.

[0157] The parameters of the dual-loop finite-time adaptive controller are set as follows: λ = 0.2, k0 = 5, k1 = 100, k2 = 100, α = 0.6, β = 0.6, λ1 = 1.4, λ2 = 2.2, μ1 = 0.3, μ2 = 1. l(x) = [5 3 1] T .

[0158] in, Figures 3-5The attitude angle response curve, attitude angle error and angular rate dynamic characteristics are shown, and the convergence time meets the fixed convergence time upper limit theoretical value T = 36.7s, Figures 6-7 The rocket three-axis position and velocity response curves show that under the conditions of considering wind disturbance, structural disturbance, engine buffeting and position loop wind resistance, the system can realize vertical stable landing control at a height of 1000m. It is shown that the double-loop finite-time controller can effectively realize high-precision attitude tracking control of the face-symmetrical launch vehicle under strong aerodynamic disturbance.

[0159] Figures 8-10 The face-symmetrical launch vehicle Monte Carlo simulation curve under system parameter deviation is shown. Within the simulated parameter uncertainty of ±20%, the standard deviation of roll angle, yaw angle and pitch angle tracking error is 0.0037deg, 0.3730deg and 0.0017deg respectively, and the pitch angle, yaw angle and roll angle tracking error can reach 10 -4 order of magnitude in 1000 simulations. Therefore, within the allowable tracking error range, the double-loop finite-time convergence control can make the face-symmetrical rocket adapt to 20% parameter uncertainty, and no instability occurs in 1000 simulations, which has good stability and response performance.

[0160] From Figures 11-14 It can be seen that compared with PD and quartic polynomial guidance, the finite-time sliding mode has smaller attitude angle overshoot, almost no oscillation and minimum steady-state error; the angular velocity oscillation amplitude is also smaller and converges faster. After replacing the sign function with the saturation function, more reasonable and smooth rudder control instructions can be generated, thereby improving the execution reliability and hardware life of the system. It has the advantages of fast convergence, strong robustness and low jitter in nonlinear systems.

[0161] According to Figures 15-18 The position simulation curve and three-axis thrust curve show that: the PD control can realize fast convergence through parameter tuning, and when there is no thrust amplitude limiting, the three-axis output can reach 10 6 order of magnitude, far exceeding the physical saturation limit of the execution mechanism. After forced limiting, the mechanical wear of the engine will be aggravated, and the system reliability will be reduced. The quartic polynomial guidance thrust instruction is naturally constrained within the feasible region without external limiting, which can avoid the saturation of the execution mechanism; and it supports explicit terminal constraints (such as attitude angle θ(t f ) = θ d , angular velocity ), which is suitable for accurate terminal guidance; but the convergence speed is limited by the polynomial order, there is phase lag, which leads to steady-state error, and high-order derivative discontinuity may cause slight chattering. In comparison, the finite-time sliding mode meets the requirements of fast convergence speed, smooth thrust and high precision.

[0162] The double-loop finite time converging face-symmetrical launch vehicle adaptive control method of the embodiment of the application is aimed at a face-symmetrical rocket model, considers the fast maneuvering and large-angle attitude tracking requirements, designs a new type of double-loop finite time high-precision attitude tracking controller for the face-symmetrical launch vehicle dynamic landing section, and can effectively realize high-precision attitude tracking and stable landing control of the face-symmetrical launch vehicle. First, relying on the fixed time sliding mode control theory, a high-performance attitude controller is designed, and an adaptive disturbance observer is combined to realize accurate attitude adaptive tracking. Then, based on the finite time theory, a position loop control scheme is constructed, through the finite time collaborative design of the position loop and the attitude loop, a six-degree-of-freedom coupling control framework is constructed, the dynamic response characteristics of the position and the attitude are optimized, and the system stability in the fast maneuvering process is ensured. Finally, under the conditions of strong aerodynamic disturbance and parameter uncertainty, the reliability and control performance of the designed double-loop finite time controller are systematically evaluated, and the robustness and superiority of the double-loop finite time controller under the conditions of complex external disturbance and parameter uncertainty are verified. The application can process the high-precision attitude tracking control problem of the face-symmetrical rocket, improve the response performance of the rocket multivariable system, and improve the flight reliability.

[0163] In the description of the present specification, the description referring to the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples without contradiction.

[0164] In addition, the terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "N" is at least two, for example, two, three, etc., unless otherwise specifically limited.

[0165] Any processes or methods described in the flowcharts or otherwise described herein can be understood as representing code modules, segments, or portions of code which include one or more executable instructions for implementing specific logic functions (or steps) of the processes, and the preferred embodiments of the application include additional implementations in which the functions are performed in a substantially simultaneous manner, or in a reverse order of presentation, or in a different order, or with additional functionality, as will be understood by those skilled in the art.

Claims

1. A bi-circular finite time converging plane-symmetry launch vehicle adaptive control method, characterized in that, The method comprises the following steps: Step 1, considering the uncertainty and external and internal disturbances, an attitude control model of a face-symmetrical launch vehicle in a dynamic landing phase is established; Step 2, an adaptive disturbance observer is designed to estimate the compound disturbance composed of structural disturbance, wind disturbance and engine vibration disturbance in real time, so that the adaptive estimation of parameters is realized; Step 3, based on the fixed-time sliding mode control theory, a high-performance attitude controller is designed in combination with the adaptive disturbance observer, so that precise high-speed large attitude maneuver tracking is realized; Step 4, in a landing coordinate system, a rigid body motion dynamics model is constructed according to a rigid body model centroid translational dynamics equation of the launch vehicle in the dynamic landing phase; Step 5, a position loop control scheme is constructed based on the finite time theory, and a six-degree-of-freedom coupling control framework is constructed through the finite time collaborative design of the position loop and the attitude loop; Step 6, based on the six-degree-of-freedom coupling control framework, a parameter setting law involved in the control law is given; Step 7, considering the actual constraint conditions, the adaptive control algorithm of the face-symmetrical launch vehicle is verified.

2. The method of claim 1, wherein, Step 1 specifically comprises: The attitude control model of the face-symmetrical launch vehicle in the dynamic landing phase is: where, are the roll, yaw and pitch angles under landing coordinates; ω = [ω x ω y ω z ] T are the three-axis attitude angular velocities under the arrow coordinates; are the equivalent swing angle commands of the rocket pitch, yaw and roll channels, ΔD is the total disturbance, including structural disturbance, wind disturbance and engine vibration disturbance; matrices R, Ω and E are respectively: where and are the structural interference moments; a w and β w are the additional wind attack and sideslip angles, respectively where W is the wind speed; A w is the wind direction angle; A0is the launch azimuth angle; V is the mass center velocity; Θ is the velocity inclination angle; The state vector is taken as Control variable Wherein, γ c is a reference roll angle, ψ c is a reference yaw angle, is a reference yaw angle, is a roll angle first derivative, is a reference roll angle first derivative, is a yaw angle first derivative, is a reference yaw angle first derivative, is a pitch angle first derivative, is a reference pitch angle first derivative, the attitude control model is converted into the following form: wherein is the second derivative of the reference attitude angle, g(x) = R · E, and assume that d(t) is bounded, i.e. there exists a constant δ > 0 such that |d(t)| ≤ δ for all t ≥ 0.

3. The method of claim 2, wherein, In step 2, the adaptive disturbance observer designed is: wherein is the disturbance estimate, l(x) is the observer gain matrix, l(x) = [l1l2l3] T , l1, l2, l3 > 0, is the designed auxiliary function; An interference estimate based on an output of an adaptive interference observer An adaptive adjustment of a time-varying control parameter k(t), the control parameter k(t) being given by the form Wherein, k0>0 is a reference gain, and λ2>0 is a controller parameter.

4. The method of claim 3, wherein, In step 3, based on the fixed-time sliding mode control theory, the following sliding mode surface s(x) is designed for the transformed model: wherein the sliding surface parameters λ1>0, μ1>0, β>0, such that has the property of terminal sliding mode, i.e. the sliding surface guarantees that the system states x1and x2converge to zero in a fixed time; The saturation function expression is as follows: The designed controller is: wherein Adapted from step 2, μ2>0 is a controller parameter, λ2is a controller parameter; - g -1 (x) f(x) belongs to the error dynamic compensation term, used to compensate the known dynamics of the system; is a nonlinear reaching law, where, guarantees the reaching speed; the exponential term makes the reaching law exhibit different convergence characteristics at different stages, i.e., the exponent is large when far away from the sliding surface and gradually decreases when approaching the sliding surface; is an additional non-linear compensation term; The stable time satisfies: wherein is a stabilization time calculation parameter.

5. The method of claim 1, wherein, In step 4, the rigid body motion dynamics model is: where r = [r x r y r z ] T is the position vector of the rocket mass center relative to the landing coordinate system O-x4y4z4, decomposed along the landing coordinate system O-x4y4z4; v = [v x v y v z ] T is the velocity vector of the rocket mass center relative to the landing coordinate system O-x4y4z4, decomposed along the landing coordinate system O-x4y4z4; g is the gravitational acceleration vector, expressed in the landing coordinate system O-x4y4z4 as wherein g r is the component of the earth's gravitational acceleration along the geocentric radius vector; g ωe is the component of the earth's gravitational acceleration along the direction of the earth's rotation; (x, y, z) are the coordinates of the rocket in the landing coordinate system; R 0x , R 0y , R 0z are the components of the geocentric radius vector of the origin of the landing coordinate system in the landing coordinate system; ω ex , ω ey , ω ez are the components of the angular velocity of the earth's rotation in the landing coordinate system; T is the thrust size of the rocket along the rocket body axis in the rocket body coordinate system, and the thrust vector T is expressed in the landing coordinate system as: wherein, is the rotation matrix from the landing coordinate system to the vehicle coordinate system; a D The air resistance acceleration vector of the rocket in the landing coordinate system is: where C D is the aerodynamic axial force coefficient; S ref is the rocket aerodynamic reference area; and p is the atmospheric density.

6. The method of claim 1, wherein, In step 5, based on the finite time theory, a position loop control scheme is constructed, and the sliding mode surface is defined as: s = e v + λe where e = r - r d , are position and velocity errors, respectively, r d is the desired position, and v d is the desired velocity, λ > 0; The control law is designed as: u = -k1|s| α • sat(s) - k2s Wherein, k1>0 and k2>0; The system stable time satisfies: where c = k1-2 (1+α) / 2 , 7. The method of claim 1, wherein, In step 6, the parameter tuning law involved in the attitude loop control law is: the feedback linear compensation term -g -1 (x)f(x) is the dynamic compensation term known to the system, which does not involve parameter tuning; the nonlinear reaching law part is where k satisfies δ is the upper bound of system disturbance, which is adaptively adjusted by the disturbance observer; λ2 controls the steepness of the exponential function, the greater the closer to the sliding mode surface, the faster the convergence; μ2 adjusts the smoothness of the exponential, which reduces the chattering when approaching the sliding mode surface; is the adaptive exponential term, which automatically adjusts the strength of the control input change with the error size; the nonlinear compensation part is where λ1 controls the nonlinear strength of the sliding mode surface, enhances the control ability near the sliding mode surface; μ1 reduces the explosion of the compensation term in extreme cases; β adjusts the overall strength of the nonlinear compensation part, from 0.1 to 2. The parameter setting law involved in the position loop control law is: λ determines the weight of the position error and the speed error in the error dynamic, controls the response speed of the sliding mode surface, and is set to 1-10; α is a nonlinear power term, determines the convergence speed and controls the chattering to a certain extent, and is set to 0-1; k1 is a nonlinear approaching term gain, mainly controls the convergence speed, the larger the value is, the greater the approaching speed is, and the control input amplitude is also increased; k2 is a linear feedback gain; in the parameter setting process, the following order is followed: firstly, fix α and λ, and increase k1 until the desired convergence speed is met, and then adjust k2 to suppress overshoot and oscillation.

8. The method of claim 1, wherein, Step 7 specifically comprises: considering the actual constraint conditions of strong aerodynamic disturbance and parameter uncertainty, verifying the robustness and superiority of the adaptive control algorithm of the face-symmetrical launch vehicle under the conditions of complex external disturbance and parameter uncertainty.

Images